National Academies Press: OpenBook

Traffic Signal Control Strategies for Pedestrians and Bicyclists (2022)

Chapter: Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic

« Previous: Chapter 5 - Introduction to Treatments
Page 31
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 31
Page 32
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 32
Page 33
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 33
Page 34
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 34
Page 35
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 35
Page 36
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 36
Page 37
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 37
Page 38
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 38
Page 39
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 39
Page 40
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 40
Page 41
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 41
Page 42
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 42
Page 43
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 43
Page 44
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 44
Page 45
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 45
Page 46
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 46
Page 47
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 47
Page 48
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 48
Page 49
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 49
Page 50
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 50
Page 51
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 51
Page 52
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 52
Page 53
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 53
Page 54
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 54
Page 55
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 55
Page 56
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 56
Page 57
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 57
Page 58
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 58
Page 59
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 59
Page 60
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 60
Page 61
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 61
Page 62
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 62
Page 63
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 63
Page 64
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 64
Page 65
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 65
Page 66
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 66
Page 67
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 67
Page 68
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 68
Page 69
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 69
Page 70
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 70
Page 71
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 71
Page 72
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 72
Page 73
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 73
Page 74
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 74
Page 75
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 75
Page 76
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 76
Page 77
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 77
Page 78
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 78
Page 79
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 79
Page 80
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 80
Page 81
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 81
Page 82
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 82
Page 83
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 83
Page 84
Suggested Citation:"Chapter 6 - Treatments that Reduce or Eliminate Conflicts with Turning Traffic." National Academies of Sciences, Engineering, and Medicine. 2022. Traffic Signal Control Strategies for Pedestrians and Bicyclists. Washington, DC: The National Academies Press. doi: 10.17226/26491.
×
Page 84

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

31   This chapter describes the following nine treatments that address conflicts with turning traffic: C H A P T E R 6 Treatments that Reduce or Eliminate Conflicts with Turning Traffic Primary Function Section Treatment Name Separation from left- turning traffic 6.1 Protected-Only Left Turns to Address Non-motorized User Conflicts Separation from right- and left-turning traffic 6.2 Concurrent-Protected Crossings 6.3 Exclusive Pedestrian and Bicycle Phases 6.4 Channelized Right Turns/Delta Islands Partial separation from right-turning traffic 6.5 Leading Pedestrian Intervals 6.6 Delayed Turn/Leading Through Intervals 6.7 Pedestrian Overlaps with Leading Pedestrian Intervals and Vehicular Holds Preventing turns on red 6.8 No Turn on Red Encouraging turning traffic to yield 6.9 Flashing Pedestrian and Bicycle Crossing Warnings Turning vehicles are one of the greatest hazards facing pedestrians and bicycles at inter- sections. They have variously been estimated to represent 25%–50% of pedestrian crashes at intersections (Lord et al., 1998). Bicyclists face the same hazard and are particularly vulnerable to right-turning vehicles. Conflicts with turning vehicles also create discomfort when pedestrians or cyclists have to compete for right-of-way with turning vehicles that do not readily yield. Most of the treatments in this chapter address this safety issue by separating pedestrians and bicycles from turning vehicles in time. The essence of traffic signal control is to separate con- flicting traffic movements into distinct phases; however, eliminating all conflicts by protecting all movements can cause large delays. This often leads to agencies allowing turn conflicts with pedestrians and bicycles, which can create safety issues due to their vulnerability. Protected-Only Left Turns to Address Non-motorized User Conflicts (Section 6.1). This treatment examines whether left turns should be protected-only—that is, completely separated in time from conflicting traffic, including crossing pedestrians and bicycles. It reveals—perhaps more than any other treatment—a large difference in practice between North America and bicycle-friendly countries in Europe, where left turns across multilane roads are nearly always protected-only. The next three treatments separate pedestrians and bicycles from right-turning as well as left-turning traffic: Concurrent-Protected Crossings (Section 6.2). Both right turns and left turns are given their own distinct phases, controlled by turn arrows, while pedestrians and bicycles cross

32 Traffic Signal Control Strategies for Pedestrians and Bicyclists concurrently with parallel through traffic. The main drawback to this treatment is that it requires an exclusive right-turn lane as well as an exclusive left-turn lane. Exclusive Pedestrian and Bicycle Phases (Section 6.3). There is one phase in the cycle for all pedestrian movements; that phase may or may not serve bicycles as well. The main draw- back is its negative impact on traffic capacity, as it can force signal cycles to be long and lead to high delay for all users, including pedestrians. Channelized Right Turns/Delta Islands (Section 6.4). Conflicts between pedestrians and right-turning traffic are removed from the main crossing. However, dealing with crossings to and from the delta islands remains a challenge. If crossings are not signalized, other fac- tors must ensure that those crossings are safe; if crossings are signalized, they must be set up as multistage crossings, which can entail large pedestrian delay unless carefully timed. The treatments in Sections 6.5–6.7 involve partial protection from right turns—that is, pre- venting right turns during an initial part of the crossing phase. Their descriptions include guid- ance on when full protection, partial protection, or no protection from right turns might be appropriate. Leading Pedestrian Intervals (Section 6.5). At the start of a vehicular phase, all traffic is held for a short time while pedestrians—and bicycles in certain cities—get a head start, allowing them to establish their priority in the crosswalk before turning traffic is released. Delayed Turn/Leading Through Intervals (Section 6.6). At the start of a vehicular phase, turning traffic is held for a short time (but typically longer than a leading pedestrian interval [LPI]) while through traffic and pedestrians—plus bicycles for certain cities—get a head start. This allows them to establish their priority in the crosswalk before turning traffic is released. Pedestrian Overlaps with Leading Pedestrian Intervals and Vehicular Holds (Section 6.7). The pedestrian phases of intersecting streets are allowed to overlap during an LPI or other short vehicular-hold interval, which can enable longer Walk intervals and make it possible to introduce LPIs with less capacity or cycle-length impact. No Turn on Red (Section 6.8). This well-known treatment supplements several other treatments described in this guidebook, such as LPI, by giving pedestrians a short interval free of turning conflicts. Flashing Pedestrian and Bicycle Crossing Warnings (Section 6.9). This treatment aims to mitigate turn conflicts by displaying flashing warnings to approaching motorists during phases with permitted-turn conflicts. Warning signs discussed include flashing yellow arrow (FYA), used for this purpose in several U.S. cities, and a flashing pictogram used in Amsterdam, Netherlands. Bibliography Lord, D., Smiley, A., & Haroun, A. (1998). Pedestrian Accidents with Left-Turning Traffic at Signalized Inter- sections: Characteristics, Human Factors, and Unconsidered Issues. In 77th Annual Transportation Research Board Meeting, Washington, DC. Available at https://safety.fhwa.dot.gov/ped_bike/docs/00674.pdf. 6.1 Protected-Only Left Turns to Address Non-motorized User Conflicts 6.1.1 Basic Description 6.1.1.1 Alternative Names None.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 33   6.1.1.2 Description and Objective Left-turning traffic is perhaps the greatest hazard that pedestrians and cyclists face at signal- ized intersections. A review of safety studies found that the proportion of pedestrian crashes at intersections that involved a left-turning vehicle was between 17% and 32% (Lord et al., 1998). In Cambridge, MA, 19% of all bicycle crashes (including crashes away from intersections) were with left-turning vehicles (City of Cambridge, 2014). In New York City, out of 859 pedestrian and bicycle fatalities in a 5-year period ending in 2014, 108 were killed by a left-turning vehicle (NYC Department of Transportation [NYC DOT], 2016). Left turns are also a leading cause of vehicle–vehicle injury crashes. Protected-only left-turn phasing separates crossing pedestrians and bicycles from left- turning vehicles in time. “Protected-only” means that left turns are allowed only during a pro- tected turn phase, in which a green arrow is displayed to left-turning traffic and no conflicting movements—including pedestrian, bicycle, and vehicular movements—run concurrently. This differs from permitted and protected-permitted left turns, in which a circular green indicates that left-turning traffic may advance after yielding to conflicting movements. The objective of making left turns protected-only is to improve safety for pedestrians and bicycles as well as for vehicles. 6.1.1.3 Variations Not applicable for this treatment. 6.1.1.4 Operating Context Protected-only left turns might be appropriate for: • Multilane roads; • Left turns across a two-way bike path running along a road; • Intersections with limited visibility between approaching bicycles and left-turning vehicles (visibility may be limited by parking, trees, etc.); • Skew intersections that allow high-speed left turns; • High volume of left turns; and • High-speed roads. Left Turn across a Multilane Road. Protected-only left turns are safer than permitted left turns at signalized intersections, especially on multilane roads. At urban signalized intersections, the crash modification factor (CMF) for changing left-turn phasing to protected-only is 0.01 for left-turn crashes, which means it virtually eliminates left-turn crashes (Highway Safety Manual, 2010). At the same time, protected phasing can increase the frequency of rear-end crashes, making its CMF 0.94 when all crashes are considered. But since head-on and angle crashes associated with left turns tend to be far more severe than rear-end crashes, there is still a substantial safety benefit associated with protected-only phasing. Permitted left turns across multiple lanes of oncoming traffic carry a particularly large col- lision risk. One comprehensive review of the literature found that with protected-permitted phasing, crash rates per left-turning vehicle were 3.2 times greater on roads where the left turn crosses two through lanes of opposing traffic versus a single lane (Hauer, 2003). A study of 200 urban intersections in Kentucky found that left-turn crash risk rises faster than exponen- tially with the number of opposing through lanes, even for a fixed volume of opposing traffic (Amiridis et al., 2017). Based on the model from the study, if there are 140 left turns per hour and opposing through volume is 700 vehicles per hour, adding a second opposing lane increases crash risk by a factor of 3.7 and adding a third lane increases it by a factor of 32.

34 Traffic Signal Control Strategies for Pedestrians and Bicyclists For crossing bicycles and pedestrians, risks involved with permitted left turns are especially high on multilane roads. For left-turning drivers on roads with only one through lane per direc- tion, finding a gap requires less attention, and as a result, they are more likely to notice pedes- trians and bicycles before beginning to turn. On multilane roads, gaps are constantly forming and dissipating because vehicles in different lanes can have different speeds, which makes scan- ning for a gap inherently more complex. Therefore, drivers waiting to turn left tend to fixate on the road as they search for and anticipate a gap. They often start turning as soon as they find a gap in the opposing through traffic, without paying attention to pedestrians and bicycles they may encounter toward the end of their crossing maneuver. Left Turn across a Two-Way Bike Path Running Along a Road. When a two-way bike path lies alongside a road, drivers turning left across the bike path face a conflict with bicycles coming from behind. This makes permitted left turns inherently risky (Massachusetts Department of Transportation [MassDOT], 2015). Limited Visibility of Approaching Bicycles. AASHTO’s A Policy on Geometric Design of Highways and Streets, 7th Edition (Green Book) recommends that left turns at signalized inter- sections have permitted phasing only if drivers waiting to turn left have a clear sight line to the approaching vehicles that they must yield to (i.e., vehicles on course to arrive before the turning vehicle would have cleared their path). While this criterion was written with opposing motor traffic in mind, it also applies to bicycles. On a road with separated bike lanes or a sidepath, either approaching bicycles close enough to conflict with a turning vehicle should be visible or the left turn should be separated in time from the bicycle movement. The time needed for the left-turn maneuver is given by Equation 6-1: = + 0.04 (6-1)Maneuver time t Co where to = 5.5 s if the design is for a left-turning passenger car, 6.5 s for a single-unit truck, or 7.5 s for a combination truck, and C = additional distance, in feet, needed to clear the conflict zone beyond the distance needed to clear the first opposing lane (see Exhibit 6-5). (For this purpose, the Green Book’s coefficient of 0.5 s per additional lane has been converted to 0.04 s per additional foot.) In Exhibit 6-1, the approach zone is the area within which the left-turning driver must yield to any approaching bicycle; Equation 6-2 describes its length as: = × (6-2)L Bicycle design speed Maneuver time Bicycle design speed may be taken to be 14.7 to 17.6 ft/s (10 to 12 mph) on level ground and greater if bicycles approach on a downgrade. Left turns should be protected-only if visibility from the left turn lane’s waiting position to any part of the approach zone is obstructed. Exhibit 6-2 illustrates an obstructed view for the left-turn maneuver. In the photo, taken from well within the approach zone, the vehicle circled in red is actually the second vehicle in the left-turn queue; the line of sight of the first vehicle waiting to turn left is completely obstructed. Fortunately, left turns at this intersection are protected-only. Skew Intersection. Where a skew intersection angle allows left turns to be made at high speed, both the risk and severity of a crash with a crossing pedestrian, bicycle, or opposing motor vehicle is elevated. High Volume of Left Turns. High left-turn volume increases the pressure drivers feel to turn (rather than wait). It also increases the fraction of vehicles that turn left as a follower in

Treatments that Reduce or Eliminate Conicts with Turning Trafc 35   a platoon. Research done for this project found that le-turning drivers in the second or later position in a platoon were 56% less likely to yield to a crossing bicycle or pedestrian than drivers who were not immediately following another vehicle. For separated bike lanes, the MassDOT Separated Bike Lane Planning & Design Guide recommends, for example, a limiting volume of 50 vehicles per hour for le turns from a two-way street across two lanes and a one-way bike path. Assuming a 90 s cycle length, that is approximately one le-turning vehicle per cycle on average, thus avoiding platooned turns in most cycles. High-Speed Roads. On high-speed roads, vehicle–vehicle crashes involving le turns are oen deadly. 6.1.2 Applications and Expected Outcomes 6.1.2.1 National and International Use In Amsterdam and other Dutch cities, protected-only phasing is used by policy at all multi- lane intersections and wherever le turns cross a two-way bike path. Permitted le turns are Exhibit 6-1. A left-turning passenger car versus an oncoming bicycle. Implementation Source: Peter Furth. Exhibit 6-2. Obstructed view between waiting vehicle and the approach zone on a sidepath in Boston, MA. The circled car is the second vehicle in the left-turn queue.

36 Traffic Signal Control Strategies for Pedestrians and Bicyclists allowed on streets with one lane per direction; however, protected left turns are more common in practice, except from minor-street approaches that are too narrow to have left-turn lanes. In the United States, guidelines regarding the use of permitted left turns are comparatively less strict than in the Netherlands. For example, guidelines found in NCHRP Report 812: Signal Timing Manual, 2nd Edition (based on other national publications and repeated in many state guidelines) suggest using protected-only phasing only under the following conditions: a vis- ibility issue; dual left-turn lanes; crossing four or more opposing through lanes; a speed limit of 50 mph or greater; or if experience with permitted left turns has led to an excessive number of left-turn crashes, more than roughly five per year per left-turn movement. Apart from the crash experience criterion, these are conditions that relatively few intersections meet. It also stands out that these guidelines make no explicit consideration for pedestrian or bicycle use. Some American communities with high pedestrian- and bicycle-use have adopted policies that favor protected-only left turns. For example, at all intersections with multilane roads in Cambridge, the city has been converting permitted left turns to protected-only left turns to provide a safer operation for pedestrians, bicycles, and other vehicles (City of Cambridge, n.d.). Boulder, CO, follows a draft policy that calls for protected-only left turns across shared-use paths with a minimum volume of 30 bicycles per hour and across crosswalks with at least 100 pedes- trians per hour (City of Boulder, n.d.). New York City has also recently converted several inter- sections to protected-only phasing. 6.1.2.2 Benefits and Impacts Protected-only left turns improve safety for pedestrians, bicycles, and turning vehicles. A before-and-after study found that when nine intersections in New York City were converted to protected-only phasing, pedestrian–vehicle crashes fell by 28% and vehicle–vehicle crashes fell by 32%. The same study found no reduction in pedestrian–vehicle crashes after conversions to protected-permitted left turns, which were applied in Chicago, IL, and Toronto, Ontario (Goughnour et al., 2018). Research done as part of this guidebook found that on a multilane road in situations where the only conflict is between a left-turning motorist and a cyclist during a permitted-turn phase, motorists failed to yield to bicycles more than half the time. This study examined two inter- sections in Boston with left turns across three lanes of traffic and a bike path offset about 10 ft from the road. In September 2019, both intersections were converted from protected-permitted phasing to protected-only. Before conversion, 164 bicycle crossings took place during the per- mitted phase while a left-turning vehicle was waiting to turn and no opposing traffic was blocking the left turn. The behavior of left-turning vehicles during this time is summarized as follows: • In only 15 cases (9%) did the left-turning vehicle wait in the turn lane until both the opposing traffic lanes and bike path were clear. • In 103 cases (63%), the left-turning vehicle claimed the right-of-way and made its turn, forcing the bicycle to stop or slow down. • In the remaining 46 cases (28%), the left-turning vehicle began to turn and then, seeing that the cyclist was not yielding, stopped until the cyclist had cleared, blocking one or two opposite-direction travel lanes. Opposite-direction vehicles sometimes had to stop to avoid crashing broadside into these stopped vehicles. After the left-turn phase was converted to protected-only, motorist failure-to-yield events all but disappeared. (Unfortunately, some red-light running by left-turning vehicles persisted several months after the conversion.) Changing the left-turn mode from permitted or protected-permitted to protected-only increases delay for left-turning vehicles and, to a lesser extent, for bicycles and pedestrians,

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 37   whose green duration becomes shorter. These impacts can be mitigated by using shorter cycles (see Section 7.1); lagging left turns (which can offer better progression for left-turning vehicles); or reservice for left turns, which in this context means giving left turns both a leading and a lag- ging phase (see Section 7.2). Research for this guidebook modeled the two Boston intersections discussed earlier and compared user delay for a base case with arterial coordination on a 120 s cycle and protected-permitted phasing against a variety of alternatives with protected-only left turns. Exhibit 6-3 compares the base case with leading protected lefts and shows that bicycle and pedestrian delays increase modestly while average left-turn delay rises by more than 45 s. However, other options result in considerably less left-turn delay, especially reservice and run- ning free. Interestingly, for this corridor, running free results in the lowest delay for vehicles, pedestrians, and bicycles, mainly because it allows a far shorter average cycle length. 6.1.3 Considerations 6.1.3.1 Accessibility Considerations Protected crossings benefit pedestrians with disabilities who are not in a strong position to compete with motor vehicles for right-of-way or to maneuver around vehicles that do not yield. It can be difficult for individuals with vision disabilities to distinguish the protected movement from through movements, so accessible pedestrian signals (APS) can help them begin crossing at the proper time. Protected-only left turns also make driving safer and easier for young (novice) drivers, whose driving skills are still developing, and for older drivers, whose perception abilities may be reduced. 6.1.3.2 Guidance For separated bike lanes, the MassDOT Separated Bike Lane Planning & Design Guide recom- mends protected-only turns when turning volumes exceed those shown in Exhibit 6-4. It recom- mends, for example, a limiting volume of 50 vehicles per hour for left turns across two lanes and a one-way bike path. Assuming a 90 s cycle length, that is approximately one left-turning vehicle per cycle on average, thus avoiding platooned turns in most cycles (the effect of platooned turns on yielding and safety was discussed previously in the High Volume of Left Turns section). 6.1.3.3 Relationships to Relevant Treatments Converting a left-turn phase to protected-only can create challenges that can be addressed by applying other treatments at the same time. • If a left-turn bay is short and might spill back to block a through lane, reservice for the left- turn phase (see Section 7.2) can be considered. Pedestrian Delay (s) Bicycle Delay (s) Left-Turn Delay (s) Intersection Vehicle Delay (s) Average Cycle Length (s) Protected-Permitted (Base Case) 25 18 27 40 120 Protected-Only (Leading Left) 27 32 73 58 120 Protected-Only (Lagging Left) 27 32 67 42 120 Protected-Only with Left-Turn Reservice 27 26 34 51 120 Protected-Only, Leading Left, Running Free 19 15 52 32 70 Exhibit 6-3. Delay impacts of left-turn phasing alternatives, Columbus Avenue and Heath Street, Boston, morning peak hour.

38 Traffic Signal Control Strategies for Pedestrians and Bicyclists • If very heavy turning demand results in a short green window for bicycles, consider small- zone coordination to limit bicycle delay (see Section 9.2). • An FYA operation can be used to run protected-only at certain times of day or when pedes- trians are detected (see Section 6.9). 6.1.4 Implementation Support 6.1.4.1 Equipment Needs and Features Changing the left-turn phase to protected-only may require additional signal heads, and adding signal heads may require additional mounting poles or mast arms. 6.1.4.2 Phasing and Timing In coordinated corridors with protected-only left turns, lagging left turns often lead to less delay for left-turning vehicles because signals are typically timed for the main platoon to arrive during the green interval. With lagging left turns, most of the vehicles arriving in the main pla- toon are served toward the end of their arrival phase, while with leading left turns, they have to wait for the next cycle. 6.1.4.3 Signage and Striping Not applicable for this treatment. 6.1.4.4 Geometric Elements Protected-only left turns typically require an exclusive left-turn lane. Bibliography American Association of State Highway and Transportation Officials. (2011). A Policy on Geometric Design of Highways and Streets, 7th Edition (Green Book). Washington, DC. Amiridis, K., Stamatiadis, N., & Kirk, A. (2017). Safety-Based Left-Turn Phasing Decisions for Signalized Inter- sections. Transportation Research Record: Journal of the Transportation Research Board, 2619(1), 13–19. City of Boulder, CO. (n.d.). [Website]. Retrieved April 28, 2020, from https://bouldercolorado.gov/transportation/ traffic-signals City of Cambridge, MA. (n.d.). [Website]. Retrieved April 28, 2020, from https://www.cambridgema.gov/traffic/ engineeringplanning/trafficsignals City of Cambridge, MA. (2014). Bicycle Crash Fact Sheet. https://www.cambridgema.gov/-/media/Files/CDD/ Transportation/Bike/bicyclesafetyfacts_final_20140609.pdf Goughnour, E., Carter, D., Lyon, C., Persaud, B., Lan, B., Chun, P., Hamilton, I., and Signor, K. (2018). Safety Evaluation of Protected Left-Turn Phasing and Leading Pedestrian Intervals on Pedestrian Safety. Report FHWA-HRT-18-044, U.S. DOT. Separated Bike Lane Operation Motor Vehicles per Hour Turning across Separated Bike Lane Two-Way Street One-Way Street Right Turn Left Turn across Two Lanes Right or Left Turn One-way 150 100 50 150 Two-way 100 50 0 100 Source: MassDOT (2015). Left Turn across One Lane Exhibit 6-4. Turning volume criteria for protected-only left and right turns.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 39   Hauer, E. (2003, January). Left-Turn Protection, Safety, Delay and Guidelines: A Literature Review. University of Toronto. Highway Safety Manual. (2010). AASHTO, Washington, DC. Lord, D., Smiley, A., & Haroun, A. (1998). Pedestrian Accidents with Left-Turning Traffic at Signalized Intersec- tions: Characteristics, Human Factors, and Unconsidered Issues. In 77th Annual Transportation Research Board Meeting, Washington, DC. https://safety.fhwa.dot.gov/ped_bike/docs/00674.pdf. Massachusetts Department of Transportation. (2015). Separated Bike Lane Planning & Design Guide. NYC Department of Transportation. (2016, August). Don’t Cut Corners: Left Turn Pedestrian & Bicyclist Crash Study. Urbanik, T., Tanaka, A., Lozner, B., Lindstrom, E., Lee, K., Quayle, S., Beaird, S., Tsoi, S., Ryus, P., Gettman, D., Sunkari, S., Balke, K., & Bullock, D. (2015). NCHRP Report 812: Signal Timing Manual, 2nd Edition. Trans- portation Research Board, Washington, DC. 6.2 Concurrent-Protected Crossings 6.2.1 Basic Description 6.2.1.1 Alternative Names Protected right turn; left-turn overlap; split through phase. 6.2.1.2 Description and Objective Pedestrian and bicycle crossings are concurrent with parallel through traffic, yet they can be separated in time from both right and left turns by providing turning movements with distinct phases (Exhibit 6-5). The objective is to improve safety by separating pedestrians and bicycles from conflicts with turning vehicles. Protecting pedestrian/bicycle crossings from left turns is discussed as its own treatment in Section 6.1. The distinction of this treatment is that crossing phases are also protected from right turns. This treatment applies to protecting crossings from left turns from one-way streets as well. Exhibit 6-5. Concurrent-protected crossings.

40 Traffic Signal Control Strategies for Pedestrians and Bicyclists 6.2.1.3 Variations Variations to this treatment arise from how the right-turn movement fits into the phasing plan. Wherever a right-turn movement has a parallel left-turn phase, it is efficient to run those turns concurrently. This scheme is illustrated in Exhibit 6-6; for example, southbound right is parallel to eastbound left, so the two movements can run together during Phase 01 (similarly, westbound right and southbound left movements can run together during Phase 03). In the controller, the right-turn phase can be programmed as an “overlap” that times concurrently with a left-turn phase. If the right-turn movement runs only during a left-turn phase, as shown in Exhibit 6-6, it is a simple overlap. More complex overlaps can also be programmed in which the right-turn phase also runs during part of the through movement, as explained later. In some situations, there is no left-turn phase parallel to the right-turn movement, such as on a one-way street or where the cross street does not have a left-turn phase. In such a case, a phasing scheme called split through phase can be applied. Shown in Exhibit 6-7 is a typical phasing plan used in New York City along a one-way avenue with a protected bike lane on the left side of the street. The through movement for the avenue—which in this example runs northbound—is split into two phases, with Phase 01 serving the left-side protected bike lane crossing and Phase 02 serving left-turning vehicles. 6.2.1.4 Operating Context This treatment is useful whenever certain conditions make it more desirable to separate cross- ings in time from right turns as well as left turns. Such conditions include heavy right-turn volumes, intersection geometry that allows high-speed right turns, and bicycle crossings with limited visibility. This treatment can also be useful when pedestrian crossings are so heavy that right-turn flow would be blocked unless right turns are given a distinct phase separated from crossing pedes- trians, which might occur in a downtown area or near a major transit station. Exhibit 6-6. Right turns served with simple left-turn overlaps in a single-ring phasing plan. Exhibit 6-7. Split through phase as applied along one-way avenues in New York City.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 41   This treatment also requires an exclusive right-turn lane. At intersections without exclusive right-turn lanes, it may be possible to create an exclusive right-turn lane by widening an inter- section approach, converting a parking lane into a turn lane, or converting a shared through- right lane into an exclusive right-turn lane and adjusting signal timing accordingly. 6.2.2 Applications and Expected Outcomes 6.2.2.1 National and International Use In Dutch cities, where protected bike lanes are common, it is typical to have concurrent- protected crossings. A road with only one travel lane per direction that widens to three lanes as it approaches an intersection is common: one lane each for left turns, through vehicles, and right turns. This allows the bicycle and pedestrian crossings to be given phases that are separated in time from those of the right turns and left turns. In the U.S., concurrent-protected crossings using left-turn overlaps and split through phases are well-established techniques, though not as commonly applied. Application has grown as cities including Portland, OR; Long Beach, CA; New York; Boston; and Cambridge, MA, have used this treatment over the last 15 years to create safer bicycle and pedestrian crossings. The experiences and policies of Amsterdam; New York City; and the island of Montreal, Québec, are instructive for understanding the trade-off between protection and delay when protecting bicycles from right-turn conflicts. In Amsterdam, concurrent-protected bicycle and pedestrian crossings became the norm in the 1970s. This led to long signal cycles and, as an unintended consequence, long delays for bicycles (and for right-turning motorists). Complaints and high rates of noncompliance led to a change in policy in the early 1980s. Since then, right- turn conflicts have become permitted at most intersections, although concurrent-protected phasing is still used on approaches with high right-turn volumes or high right-turn speeds and anywhere that right turns can be served with a simple left-turn overlap (Linders, 2013). New York City’s policy regarding protected crossings evolved similarly to Amsterdam’s. Caution over implementing the nation’s first parking-protected cycle tracks in 2007 resulted in concurrent-protected crossings at nearly all intersections with protected bike lanes. After several years, high cyclist noncompliance at intersections with low turn volumes led officials to recognize that at such locations, conflicts with permitted left turns were not a significant hazard. As a result, many concurrent-protected crossings were converted to the delayed-turn treatment (see Section 6.6) in which the bicycle crossing is protected-only during an initial interval (e.g., 10 s or 15 s), after which conflicting left turns (from a one-way street) are allowed and governed by an FYA (see Section 6.9) to alert drivers of a potential conflict. New York City’s current policy prefers concurrent-protected crossings only where there is high volume or high-speed turn conflicts (provided a turn lane is available or can be created). Without these conditions, if a turn lane can be provided then delayed turn is preferred; where no turn lane can be created, an LPI is preferred (see Section 6.5) (D. Nguyen, personal communication, October 10, 2019). Montreal approached the protection-delay trade-off from the other direction. Montreal has long used the delayed-turn treatment (see Section 6.6) at intersections throughout its down- town. When a downtown two-way bicycle path was created in 2007 along the left side of Avenue de Maisonneuve, a one-way street, delayed turn was applied. Bicycles had a protected crossing for the first 9 s but were concurrent with through traffic after that, and left turns across the bicycle path were permitted during the rest of the through phase. However, over the years, bicycle volume increased, and motorists could not find enough safe gaps to turn left, particu- larly since the bike path is two-way; this led to unsafe turning behaviors, such as drivers forcing a gap. In the face of mounting complaints, signals were changed in 2019 to make crossings

42 Traffic Signal Control Strategies for Pedestrians and Bicyclists concurrent-protected using split through phasing. The left lane was converted to an exclusive left-turn lane, leaving only one lane for through traffic. Initial implementation led to complaints of large increases in cyclist delay, leading the city to commit to improving bicycle progression (see Section 9.2). 6.2.2.2 Benefits and Impacts While exclusive pedestrian and bicycle phases (see Section 6.3) and concurrent-protected crossings both provide fully protected crossings, concurrent-protected phasing usually provides more vehicular capacity, allows shorter cycles, and results in less delay for all users, especially where right turns can be served using left-turn overlaps. A simulation study of a four-leg inter- section in Boston currently operating with exclusive pedestrian and bicycle phases found that by using left-turn overlaps, concurrent-protected phasing would reduce the needed cycle length from 135 s to 93 s while lowering average delay by 17 s for vehicles and by 22 s for pedestrians (Furth et al., 2014). Concurrent-protected crossings eliminate turn conflicts, which makes them safer than cross- ings with permitted right-turn conflicts. Other impacts include increased delay to vehicles (especially those turning right), increased delay to bicycles and pedestrians, and an enlarged intersection footprint to facilitate a right-turn lane. In many cases, however, concurrent-pro- tected crossings can be implemented with no footprint impact. When the junction of Broadway and Galileo Galilei Way, Cambridge, was converted in 2017 to concurrent-protected phasing, two of the approaches already had right-turn lanes; on the other two approaches, the rightmost through lanes were converted to exclusive right-turn lanes. The through phases still had suf- ficient capacity, in spite of having one fewer lane, because right-turning traffic and pedestrian interference were removed. Cycle length was left unchanged, and the level of service for vehicles was also unchanged. However, the treatment has been called a “night and day” improvement for pedestrians because they can now cross without danger from right-turning cars (P. Baxter & C. Seiderman, personal communication, December 19, 2018). There is almost no delay impact to pedestrians or bicycles when right turns can be served using simple overlap with a left-turn phase; if this is not possible, concurrent-protected cross- ings can increase pedestrian and bicycle delays by limiting their crossing phase to only part of the through phase. Delay impacts will vary from site to site and are difficult to generalize. One study, based on the junction of NW Broadway and NW Lovejoy Street in Portland, found that providing a protected crossing increased average cyclist delay by 4 s to 16 s, depending on traffic volumes (Furth et al., 2014). 6.2.3 Considerations 6.2.3.1 Accessibility Considerations The surge of traffic by right-turning vehicles using a protected right-turn phase may be incor- rectly interpreted as the beginning of the parallel through-traffic phase and the simultaneous onset of the Walk interval by individuals who cannot see the Walk signal. The complexity and potential changes in signal phasing for each cycle also can be confusing to such pedestrians. Provide an APS to provide instruction for these users. 6.2.3.2 Guidance There is no national guideline for an acceptable volume of permitted right-turn conflicts. Both the State of Massachusetts and Boston recommend protected pedestrian crossings where con- current right turns would exceed 250 vehicles per hour (which is approximately seven vehicles per cycle, since most intersections have a cycle length of 100 s). Dutch guidelines set the limit at

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 43   150 vehicles per hour for one-way cycle tracks and recommend that two-way cycle tracks avoid all permitted conflicts (an increasing number of exceptions to this latter rule can be found in the Netherlands). 6.2.3.3 Relationships to Relevant Treatments Protected-only left turns to address non-motorized user conflicts (see Section 6.1) and no turn on red (NTOR) (see Section 6.8) are necessary parts of this treatment. Exclusive pedestrian and bicycle phases (see Section 6.3) are an alternative treatment that provides fully protected pedestrian crossings. Delayed turn/leading through intervals (see Section 6.6) are an alternative treatment that provides partially protected crossings. Channelized right turns (see Section 6.4) also provide concurrent-protected crossings, but they allow right turns to run at the same time as crossing bicycles and pedestrians because their conflict is resolved in advance of the intersection. 6.2.4 Implementation Support 6.2.4.1 Equipment Needs and Features New right-turn phases will need signal heads and support structures. 6.2.4.2 Phasing and Timing Exhibit 6-6 presents an example of how concurrent-protected crossings with left-turn over- laps can be arranged using a single ring. This is the phasing plan used in Cambridge at the junc- tion of Broadway and Galileo Galilei Way, as described earlier. It is most appropriate when the left- and right-turn movements that run concurrently have similar volumes. The same sequence can be arranged in a dual ring, offering greater flexibility to match phase lengths to demand (see Exhibit 6-8). It is the same as the standard dual ring, except that right turns run with their parallel left turn instead of their parallel through movement. Notice how any given crossing in Exhibit 6-8 (e.g., the north-side crossing, running during Phase 02) has a clear sequence: The conflicting left turn leads (Phase 01), the crossing is in the middle (Phase 02), Exhibit 6-8. Dual-ring phasing using simple left-turn overlaps to serve right turns.

44 Traffic Signal Control Strategies for Pedestrians and Bicyclists and conflicting right-turn lags (Phase 03). As with any dual-ring structure, phase sequence can be reversed so that the conflicting right turn leads and the conflicting left turn lags. Exhibit 6-9 has the same ring-barrier structure, but with right turns served using complex overlaps, providing additional flexibility to serve right-turn demands. For this example, left turns are lagging and right turns are leading. Each through movement has been subdivided into two phases; for example, westbound through is served by Phases 09 and 02, which together comprise Overlap A. During the first part of Overlap A (i.e., Phase 09), westbound right turns can run; during the latter part (Phase 02), the north-side crossing runs fully protected. West- bound right turns are served not only during Phase 09 but also during Phase 07, concurrent with a parallel left turn. The two phases serving westbound right turns comprise Overlap H. This example has four overlaps for the through movements (A, B, C, D) and four overlaps for right turns (E, F, G, H). This more complex ring structure allows phase lengths to more freely adapt to demand—in particular, it allows right turns to run longer than their parallel left turn, which can be helpful for serving streets whose dominant flow direction changes between morning and evening peaks. This ring structure, like all the others presented in this section, can be used with either pretimed or actuated control. With any of these phasing plans, bicycle delay will be minimized if the conflicting right-turn phase is actuated so any time that is not needed by right-turning vehicles can revert to the crossing phase. This is most easily accomplished by sequencing right turns to precede the pro- tected crossing (Furth et al., 2014). Where the conflicting crossing is short, the crossing phase and/or right-turn phase can appear several times within a cycle, a technique called reservice (see Section 7.2). 6.2.4.3 Signage and Striping Not applicable for this treatment. Exhibit 6-9. Using complex overlaps to create a full set of concurrent-protected crossings.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 45   6.2.4.4 Geometric Elements Concurrent-protected crossings require exclusive right-turn lanes. In some cases, a through lane can be converted into a right-turn lane while still providing sufficient through capacity. Bibliography City of Boston. (2013). Boston Complete Streets Design Guidelines. CROW. (2016). Design Manual for Bicycle Traffic. Ede, Netherlands. Furth, P. G., Koonce, P. J., Miao, Y., Peng, F., & Littman, M. (2014). Mitigating Right-Turn Conflict with Protected Yet Concurrent Phasing for Cycle Track and Pedestrian Crossings. Transportation Research Record: Journal of the Transportation Research Board, 2438(1), 81–88. Harkey, D. L., Carter, D., Bentzen, B. L., & Barlow, J. M. (2007). NCHRP Web-Only Document 117A: Accessible Pedestrian Signals: A Guide to Best Practices. Transportation Research Board, Washington, DC. Lin, P. S., Wang, Z., Guo, R., & Kourtellis, A. (2016, May). A Pilot Study on Interactions between Drivers and Pedestrian Features at Signalized Intersections—Using the SHRP2 Naturalistic Driving Study Data. In Pro- ceedings of the 11th Asia Pacific Transportation Development Conference and 29th ICTPA Annual Conference, 70–77. Linders, S. (2013). STOP! 100 jaar verkeer regelen in Amsterdam, 1912–2012. City of Amsterdam. Magder, J. (2019, November 8). Out of Sync: ‘Safer’ Downtown Traffic Lights Exasperate Montreal Cyclists. Montreal Gazette. https://montrealgazette.com/news/local-news/out-of-sync-safer-downtown-traffic- lights-exasperate-cyclists 6.3 Exclusive Pedestrian and Bicycle Phases 6.3.1 Basic Description 6.3.1.1 Alternative Names Pedestrian and/or bicycle scramble; Barnes Dance. 6.3.1.2 Description and Objective With exclusive pedestrian or bicycle phases, the pedestrian/bicycle crossings phase occurs while all vehicular movements have a red indication. This treatment aims to increase pedestrian/ bicycle safety by eliminating turn conflicts. It is also sometimes used to increase capacity for right turns when high flows of pedestrians block concurrent right turns, to increase pedestrian capacity when pedestrian volumes are high, and to enable pedestrians and/or bicycles to make diagonal crossings. 6.3.1.3 Variations Exclusive phases may or may not include diagonal crossings (Exhibit 6-10[a] and Exhibit 6-10[b], respectively). Where diagonal crossings are formally allowed, an exclusive pedestrian phase can be called a pedestrian scramble or Barnes Dance. Even when a diagonal crossing is not formally provided, many people still cross diagonally. If exclusive pedestrian phases are provided, pedestrians can also be allowed to cross concur- rently (i.e., concurrent with the parallel vehicular movement), as illustrated in Exhibit 6-10(c). This option results in high pedestrian-crossing capacity and less waiting time for pedestrians, but pedestrians are not protected from permitted turns during concurrent phases. Exclusive pedestrian phases are common; exclusive bicycle phases less so. In the U.S., exclu- sive bicycle phases typically permit only certain non-conflicting bicycle movements (e.g., a diag- onal crossing); in the Netherlands, all bike directions are allowed, which creates cross-direction conflicts that bicyclists appear to resolve informally without any safety issues. Exclusive phases shared by pedestrians and bicycles are even less common, although bicycles often run informally during exclusive pedestrian phases.

46 Traffic Signal Control Strategies for Pedestrians and Bicyclists 6.3.1.4 Operating Context There are several contexts for which exclusive pedestrian or bicycle phases might be appropriate: • Where either high-speed right turns, a high volume of right turns, or frequent right-turning trucks make concurrent crossings unsafe, and it is not possible to provide a dedicated right- turn lane; • At intersections with very high pedestrian-volumes—as might be common near a busy transit station—where concurrent crossings would conflict with right-turning traffic to the point of creating tension and/or overly restricting right-turn capacity; • At intersections with very high pedestrian-volumes—where pedestrians need the green for a large part of the cycle—by combining an exclusive pedestrian phase with concurrent cross- ings (see Exhibit 6-10[c]); • Where there is high demand for diagonal pedestrian crossings; or • To serve an important diagonal bicycle crossing, such as when a bicycle path switches from one side of the road to another. 6.3.2 Applications and Expected Outcomes 6.3.2.1 National and International Use Exclusive pedestrian phases are used widely in North American cities, most often in downtowns. While many applications formally provide for diagonal crossings, some do not. This is because diagonal crossings require a longer pedestrian clearance time. For example, downtown Denver, CO, has many intersections with exclusive pedestrian phases that originally featured diagonal Source: Kattan et al. (2009). Exhibit 6-10. Typical phase sequences with an exclusive pedestrian phase: (a) with diagonal crossings, known as pedestrian scramble or Barnes Dance; (b) without formal diagonal crossings; and (c) with pedestrians also allowed to cross concurrently.

Treatments that Reduce or Eliminate Conicts with Turning Trafc 47   crossings. However, several years ago, the city retimed its signals for a crossing speed of 3.5 /s instead of 4 /s, and diagonal crossings were formally removed to avoid having to lengthen pedestrian phases. Several intersections in downtown Toronto have exclusive pedestrian phases in addition to concurrent crossings during a parallel vehicle phase (see Exhibit 6-10[c]), resulting in high capacity and low delay for pedestrians. In downtown Washington, DC, the same treatment can be seen at the intersection of 7th Street NW and H Street NW. In Massachusetts, exclusive pedestrian phases have long been the default treatment for intersections on state highways, and they are also common at local, municipal intersections. However, they typically involve long pedestrian waiting times and poor pedestrian compliance; therefore, pedestrian advocates generally prefer concurrent crossings, except where right-turn volumes are high or turns are made at high speeds due to intersection geometry. New York City has more than 80 exclusive pedestrian phase locations, typically where skewed geometry allows for high-speed right turns, where there is a strong desire for pedestrians to cross diagonally to and from major transit stations, or where there is a high volume of turning vehicles. e city also has 386 “T-away” intersections, which are T-intersections where the cross street is one-way headed away from the intersection, making the cross-street phase a de facto exclusive pedestrian phase (NYC DOT, 2017). Exclusive bicycle phases are far less common. One type is a phase that serves a diagonal crossing—such as where a bicycle path switches from one side to another—and other bicycle movements are not allowed during that phase. Most oen, pedestrian movements are also not allowed. Portland has two such applications, including one at North Interstate Avenue and Oregon Street. At 9th Avenue North and Westlake Avenue North in Seattle, WA (see Exhibit 6-11), the diagonal bike crossing phase doubles as an exclusive pedestrian phase; however, other bike move- ments are held. The other type of exclusive bicycle phase is a bicycle scramble phase, in which bicycles in all directions get a green signal. This treatment is used at 28 intersections in Groningen, Netherlands, and at a few intersections in other Dutch cities, where it is called “All Directions Source: Kittelson & Associates, Inc. Exhibit 6-11. Diagonal bicycle crossings at 9th Avenue North and Westlake Avenue North in Seattle, Washington.

48 Traffic Signal Control Strategies for Pedestrians and Bicyclists Green.” The main impetus for applying this treatment has been to protect cyclists from con- flicts with turning traffic. Where it has been applied in Groningen, it has eliminated fatal bicycle–motor vehicle collisions; bicycle–bicycle conflicts, which are resolved without formal rules, have not been a safety problem. In some applications, pedestrians also cross during the exclusive phase. In order to minimize bicycle delay, the bicycle phase comes up twice per cycle wherever intersection capacity allows; about 25% of the Groningen intersections with bicycle scramble have two bicycle phases per cycle all day long, and at a few others, bicycles get two phases per cycle outside of peak hours (City of Groningen, 2019; J. Valkema, personal com- munication, November 20, 2019). 6.3.2.2 Benefits and Impacts Exclusive pedestrian phases have been shown to reduce collisions and conflicts involving pedestrians. One study found that at intersections in New York City where concurrent crossings were replaced with exclusive pedestrian phases, pedestrian crashes fell 50% versus a 4% decrease for a control group. At the same time, however, vehicle crashes increased 10% at the treatment site, while they decreased 12% at the control sites (Chen et al., 2014). At an intersection in Oakland, CA, that had been converted from concurrent crossings to an exclusive pedestrian phase, pedestrian–vehicle conflicts in which one party had to stop or change course unusually to prevent collision fell from 11.8 to 6.4 conflicts per 1,000 pedestrians (Bechtel et al., 2004). A similar study of two converted intersections in Calgary, Alberta, Canada, also found a signifi- cant decrease in pedestrian–vehicle conflicts (Kattan et al., 2009). Because exclusive pedestrian phases decrease the time available for vehicular movements, they can increase the necessary cycle length substantially, increasing delay for pedestrians and vehicles alike. If an exclusive phase will last 20 s, for example, that may require increasing the cycle length by 40 s or more to maintain vehicular capacity, since the vehicular phases will have longer red periods and will therefore require longer green periods. For some intersec- tions, adding an exclusive pedestrian phase will put intersections over capacity. On the other hand, there can be a countervailing effect if pedestrians create so much right-turn blockage that the saturation flow rate declines precipitously with concurrent crossings. In such a case, isolating pedestrians within a single phase can make the vehicular phases much more efficient; this reduces the negative capacity and delay impacts of exclusive pedestrian phases, particularly where right-turn blockage affects both intersecting streets. The increase in delay caused by exclusive pedestrian phases is not only a drawback in itself; it can also promote pedestrian noncompliance, diminishing the technique’s safety benefits. The Oakland and Calgary studies (Bechtel et al., 2004; Kattan et al., 2009) both found a large increase in pedestrian noncompliance, with many people crossing concurrently with the par- allel vehicular phase. Another NYC study found that applying exclusive pedestrian phases with diagonal crossings at five intersections with high pedestrian-volumes increased waiting times for all roadway users, interrupted pedestrian walking flow, and led to sidewalk overcrowding (NYC DOT, 2017). 6.3.3 Considerations 6.3.3.1 Accessibility Considerations Exclusive pedestrian phases may not be recognized by pedestrians who are visually impaired or who have low vision if APS are not installed. They will typically cross with the movement of concurrent vehicles in the absence of accessible signal information. The Manual on Uniform Traffic Control Devices (MUTCD) (2009) notes this as an issue when considering APS in Part 4, Section E.09, Part 03.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 49   6.3.3.2 Guidance The Toronto Transportation Division developed the following guidelines for implementing exclusive pedestrian phases. The treatment should be implemented only if one or more of the following conditions are satisfied (Kattan et al., 2009): • The intersection experiences a high volume of pedestrians (3,000 per hour for an 8-hour period). • There is a combination of a moderate volume of pedestrians (2,000 per hour for an 8-hour period) with high turning-vehicle volumes (30% of the total vehicular traffic). • There is moderate pedestrian volume with high pedestrian–vehicle collisions (three collisions over the past 3 years). • There is moderate pedestrian volume, and 25% of pedestrians desire to cross diagonally. • The intersection geometry is unusual (e.g., highly skewed; five or six legs). 6.3.4 Implementation Support 6.3.4.1 Equipment Needs and Features Not applicable for this treatment. 6.3.4.2 Phasing and Timing Not applicable for this treatment. 6.3.4.3 Signage and Striping Restrictions for NTOR (see Section 6.8) should be applied through either a static sign or a blank-out sign that is active during the exclusive phase. If diagonal crossings are permitted during an exclusive pedestrian phase, diagonal striping to indicate those movements is recommended, as shown in Exhibit 6-12. At intersections where bicycles are permitted to use exclusive pedestrian phases, signage may be required to inform cyclists that crossings are allowed during the exclusive phase. Source: MUTCD (2009), Figure 3B-20. Exhibit 6-12. Example of crosswalk markings for an exclusive pedestrian phase that permits diagonal crossing.

50 Traffic Signal Control Strategies for Pedestrians and Bicyclists 6.3.4.4 Geometric Elements If bicycles will be allowed to use an exclusive phase with pedestrians, it is preferable for the intersection to be configured so that the paths of bicycles and cross-direction pedestrians meet outside the crosswalks that are regulated by traffic signals. An example of this is when bicycles are in a shared-use path or in protected bike lanes offset far enough from the curb that pedes- trians have a waiting platform between the curb and the protected bike lane. Bibliography Bechtel, A. K., MacLeod, K. E., & Ragland, D. R. (2004). Pedestrian Scramble Signal in Chinatown Neighbor- hood of Oakland, California: An Evaluation. Transportation Research Record: Journal of the Transportation Research Board, 1878(1), 19–26. Bonneson, J., Sunkari, S. R., Pratt, M., & Songchitruksa, P. (2011). Traffic Signal Operations Handbook (No. FHWA/ TX-11/0-6402-P1). Research and Technology Implementation Office, Department of Transportation, Texas. Chen, L., Chen, C., & Ewing, R. (2014). The Relative Effectiveness of Signal Related Pedestrian Countermeasures at Urban Intersections—Lessons from a New York City Case Study. Transport Policy, 32, 69–78. City of Groningen. (2019, November 19). Green Light for All Cyclists. How Does It Work? https://groningen fietsstad.nl/en/green-light-for-all-cyclists-how-does-it-work/ Hediyeh, H., Sayed, T., & Zaki, M. H. (2015). The Use of Gait Parameters to Evaluate Pedestrian Behavior at Scramble Phase Signalized Intersections. Journal of Advanced Transportation, 49(4), 523–534. Kattan, L., Acharjee, S., & Tay, R. (2009). Pedestrian Scramble Operations: Pilot Study in Calgary, Alberta, Canada. Transportation Research Record: Journal of the Transportation Research Board, 2140(1), 79–84. Korve, M. J., & Niemeier, D. A. (2002). Benefit-Cost Analysis of Added Bicycle Phase at Existing Signalized Intersection. Journal of Transportation Engineering, 128(1), 40–48. Kothuri, S. M., Kading, A., Smaglik, E. J., & Sobie, C. (2017). Improving Walkability through Control Strategies at Signalized Intersections. National Institute for Transportation and Communities. Manual on Uniform Traffic Control Devices for Streets and Highways. (2009). FHWA, U.S. DOT. http://mutcd. fhwa.dot.gov/ National Association of City Transportation Officials. (2019). Don’t Give Up at the Intersection: Designing All Ages and Abilities Bicycle Crossings. NYC Department of Transportation. (2017). Walk This Way—Exclusive Pedestrian Signal Phase Treatments Study. https://www1.nyc.gov/html/dot/downloads/pdf/barnes-dance-study-sept2017.pdf Wolfe, M., Fischer, J., Deslauriers, C., Ngai, S., & Bullard, M. (2006). Bike Scramble Signal at N Interstate & Oregon. IBPI Research Digest. 6.4 Channelized Right Turns/Delta Islands 6.4.1 Basic Description 6.4.1.1 Alternative Names Pork chop islands; right-turn slip lane. 6.4.1.2 Description and Objective A channelized right-turn lane is a lane that diverges from through lanes as it reaches an inter- section, forming a delta island between them. The delta island, also called a pork chop island, serves as a pedestrian refuge and may also serve as a bicycle refuge (Exhibit 6-13). Where right-turn volumes are high, channelized right turns result in shorter main crossings that are fully protected from right turns and, by making traffic flow more efficient, can enable a shorter cycle length and/or a smaller footprint for the intersection. At the same time, channelized right turns also pose challenges to pedestrians and cyclists that can make it advantageous to eliminate them. When channelized right turns are unsignalized, replacing them with a conventional intersection layout with square corners forces right turns to be made at a lower speed. When channelized right turns are signalized, they create multistage crossings that, if not timed carefully, can entail extremely long pedestrian and cyclist delay (see

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 51   Chapter 10). The suitability of small delta islands as pedestrian and/or cyclist refuge islands is often questionable, and eliminating channelized right turns can free up space in intersection corners that can be used to make a safer crossing layout. 6.4.1.3 Variations Channelized right turns can be controlled by traffic signals or by Stop signs, and they can also be under Yield control (either signed or, since the crosswalk has priority, implicit). Where under Stop control or Yield control, the corner geometry should promote low turning speeds and yielding compliance using measures such as sharp turning radius, raised crossings, and promi- nent signs, as seen earlier in Exhibit 6-13. Where signalized, signals should be timed to provide good progression for pedestrians and bicyclists, who will have to make a multistage crossing. Delta islands can serve as a refuge for pedestrians only, for a shared-use path, or for pedes- trians and bicycles separately. A separate bike lane through a delta island can be called a pro- tected pocket lane. If the slip lane is signalized, bicycles will have a presignal that allows them to cross the slip lane. Exhibit 6-14 shows a protected pocket lane and a bicycle presignal at a Copenhagen, Denmark, intersection. 6.4.1.4 Operating Context Channelized right-turn lanes with delta islands might be appropriate where: • A moderate or heavy right-turn flow calls for a protected bicycle/pedestrian crossing. (An alternative treatment is a concurrent-protected crossing without a slip lane [Section 6.2].) • There is a sharp right turn and a need to accommodate large design vehicles. In such a case, a channelized right turn can make the crossing much shorter. Signalizing the crossing of a delta island might be appropriate where: • Right-turn volume is high. • A skew angle allows right turns to be made at high speed. • There are poor sight lines between right-turning vehicles and crossing pedestrians/bicycles. • Less restrictive measures to ensure crossing safety and compliance, such as raised crossings and prominent signs, have not been successful. Source: Google. Exhibit 6-13. Channelized right turn in Boulder (US 36 at Baseline Road) with delta island that serves as a refuge for a shared-use path.

52 Traffic Signal Control Strategies for Pedestrians and Bicyclists 6.4.2 Applications and Expected Outcomes 6.4.2.1 National and International Use In the U.S., channelized right turns are common, particularly on wide, higher speed roads. Most were not made for the benefit of pedestrians or bicyclists but rather to reduce motorist delay by increasing turning speed. Most channelized right turns are unsignalized, but signaliza- tion is not unusual. Boulder has been a leader in improving the design of channelized turns that involve a shared-use path crossing. Some U.S. cities, including Chicago, IL, have announced plans to remove all channelized right-turn lanes due to safety issues. Channelized right turns are rarely used in Dutch cities because the slip-lane crossings either become a safety problem if left unsignalized or create unacceptable delay for pedestrians and bicycles if signalized (S. Linders, personal communication, August 1, 2018). On the other hand, Copenhagen has a prominent application at an intersection that is heavily used by bicycles and pedestrians (see Exhibit 6-14). 6.4.2.2 Benefits and Impacts A study of about 400 intersections in Canada found that intersection approaches with chan- nelized right-turn lanes and those with shared through/right lanes had around 70%–80% fewer pedestrian crashes than approaches with conventional right-turn lanes (Potts et al., 2014). However, channelized right-turn lanes can also have negative consequences for bicycles and pedestrians. Where the crossings are unsignalized, they can involve high speeds and consume space that might be used to lay out safer crossings. For example, the channelized right-turn lane in Exhibit 6-15 consumes most of the available right-of-way in the intersection corner, forcing the shared-use path to abut the curving roadway with no offset. This creates a blind conflict for bicycles who have to turn 90 degrees to enter the crossing—they have to look behind them for conflicting traffic. At the same time, motorists have no warning of whether an approaching Source: Bachiochi & Furth (2016). Exhibit 6-14. Delta island in Copenhagen (Fredensbro and Øster Søgade) with bicycle presignal and protected pocket bike lane.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 53   cyclist intends to turn into the crossing. Eliminating those channelized right turns could free up enough space to create a protected intersection layout in which the shared-use path is offset several feet from the road, improving visibility between bicycles and right-turning vehicles. When protected crossings are needed, channelized right turns enable more efficient traffic flow than either concurrent-protected crossings or exclusive crossings because they allow through traffic, right-turning traffic, and the main crossing movement to all run concurrently. The shorter main crossing can also help enable a shorter cycle length (see Section 7.1). A micro- simulation study showed that the use of yield-controlled channelized right-turn lanes can reduce vehicle right-turn delay by 25% to 75% in comparison to intersection approaches with conventional right-turn lanes (Potts et al., 2014). Channelized right turns that are not signalized have little impact on pedestrian and bicycle delays. However, if they are signalized, crossings for pedestrians and bicycles become multistage, which can lead to far greater pedestrian and bicycle delays unless the crossing phases are timed to give bicycles and pedestrians good progression. Pedestrian/bicycle progression through delta islands can be especially poor when slip lanes have no dedicated right-turn signal and instead follow the through movement’s signal. Research done in developing this guidebook found that at one such intersection, a shared-use path had an average bicycle delay of 66 s, while it would be 10 s if the slip lane were unsignalized and 14 s if it were signalized in a way that offers good progression. When delay is that long, poor compliance can be expected, which can negate any safety benefit hoped for by adding signals. Source: Google. Exhibit 6-15. Channelized right turn with a shared-use path immediately next to the curb, creating a conflict with poor visibility for cyclists and insufficient opportunity for drivers and cyclists to react.

54 Traffic Signal Control Strategies for Pedestrians and Bicyclists 6.4.3 Considerations 6.4.3.1 Accessibility Considerations APS must be carefully installed and adjusted when used at channelized right turns (see MUTCD, Section 4E.13). Design of the channelizing island should consider how pedestrians who are visually impaired will approach to ensure that the APS are clear. The channelizing island may add an unsignalized pedestrian crossing to a signalized crossing if the right turn is free or yield-controlled. The accessible pushbutton should be located on the channelizing island to avoid implying that the crosswalk between the edge of the road and the island is also protected. NCHRP Research Report 834: Crossing Solutions at Roundabouts and Channelized Turn Lanes for Pedestrians with Vision Disabilities provides assessment materials to determine if channel- ized right-turn lane crossings are accessible to users with disabilities and, depending on the results of the assessments, suggested treatments (Schroeder et al., 2017). Channelization and dif- ferent materials in the non-walking area are recommended on the island to provide wayfinding direction to pedestrians. 6.4.3.2 Guidance MUTCD (2009) guidance on crossing distance is to provide crossing from the curb/edge of a shoulder to the far side of a traveled way. This can be interpreted to mean from one channelizing island to the next, as this is the distance needed for a user to cross to a safe spot. 6.4.3.3 Relationships to Relevant Treatments If channelized right turns are signalized, they create multistage crossings (see Chapter 10)— this requires evaluating pedestrian and bicycle delays (see Sections 3.4 and 3.5) and using signal timing treatments such as reservice (see Section 7.2) to create good progression for pedestrians and bicycles. Independently mounted pushbuttons are also needed where right turns are signalized (see Section 8.3). FYAs can also be used with channelized right turns to warn of conflicts with crossing pedes- trians and bicycles (see Section 6.9). 6.4.3.4 Other Considerations In most cases, the pedestrian benefit to channelizing islands (i.e., reducing main crossing dis- tance) can be achieved by reducing the curb radius and making the intersection smaller overall. This will have the additional benefit of reducing the conflict speed between turning vehicles and pedestrians. This may not be possible in cases such as highly skewed intersections or large design vehicles. In these cases, a well-designed channelizing island can be used. Drainage, paving, and snow removal should be considered when exploring the use of chan- nelized right turns. 6.4.4 Implementation Support 6.4.4.1 Equipment Needs and Features If right-turn lanes are signalized, it is preferable that they be controlled by their own signal heads (rather than following a green ball given to the through phase) and have a detector used to actuate the turn phase to minimize pedestrian delay.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 55   6.4.4.2 Phasing and Timing Where channelized right turns are signalized, it makes the pedestrian—and sometimes bicycle— crossings multistage, which can result in long delays unless signals are timed for good progres- sion for crossing pedestrians and cyclists. This is a challenge because four different streams of pedestrian movements cross any given channelized right-turn lane (i.e., people walking north- bound, southbound, eastbound, and westbound), all arriving or departing at a different time. Creating good progression requires either limiting the right turn to a short phase or providing multiple crossing phases (and therefore multiple right-turn phases) per cycle, a tactic called reservice (see Section 7.2). When a right turn and its crossing have no conflicts except with each other, they can be con- trolled as an independent intersection running free, with short, alternating phases served on demand. An example is the control planned for a channelized right-turn lane in Boston where Tremont Street turns onto Melnea Cass Boulevard. The right turn will end whenever the pedes- trian pushbutton is actuated, and it is subject to a 10 s minimum green for the right turn, guar- anteeing a short wait for pedestrians. Because the crossing phase lasts only 12 s, right-turning vehicles also experience low delay. Exhibit 6-16 shows the phasing plan used for the Copenhagen intersection shown earlier (in Exhibit 6-14), with a channelized right turn in its northeast corner. The right-turn phase is actu- ated, so it gets only the time that it needs to clear the queue; for the rest of the cycle, the crossing A. Main red phase: Bikes advance to island. B. Left turn overlap: Serve right turners. C. Early through phase: Bikes depart, right turns run until queue clears. D. Later part of through phase: Bikes pass through. Green for SouthboundRed for Southbound Exhibit 6-16. Phasing/progression plan for Copenhagen intersection for channelized right turns.

56 Traffic Signal Control Strategies for Pedestrians and Bicyclists phase runs. In addition, the timing gives bicycles two progression windows per cycle. Bicycles arriving during Interval A cross the island, wait a short time, and depart during Interval C. Bicy- cles arriving during Intervals B and C, when right-turning traffic runs, cross to the island and pass straight through during Interval D, the latter part of the parallel traffic green. This greater pedestrian-crossing time (for the main crossing) makes it impossible to provide pedestrians with two progression windows per cycle, but they still get reasonably good progression. Southbound pedestrians use the early progression window described earlier for bicycles. Northbound pedes- trians have even better progression, entering during Interval C and departing during Interval D or Interval E. 6.4.4.3 Signage and Striping Some channelizing islands are simply painted, without a physical median. These allow for larger truck movements, but they do not provide protection for pedestrians. Painted islands are not detected or recognized by pedestrians who are visually impaired, so they might cross outside the crosswalk area. If a channelized turn-lane crossing is unsignalized and will be used by bicycles, signs are needed to inform motorists to yield to bicycles (unless motorists have a Stop sign), since the presence of a crosswalk is not sufficient to establish that obligation. 6.4.4.4 Geometric Elements Delta islands must be large enough to hold the pedestrians and/or bicycles who are expected to wait on them during high-demand cycles. Where periodic demand surges occur (e.g., at schools or sports venues), demand per cycle during such a surge should be accounted and designed for in the timing strategy. According to AASHTO’s A Policy on Geometric Design of Highways and Streets, 7th Edition (Green Book), a delta island should be no less than 50 square ft in urban areas (75 square ft in rural areas); delta islands larger than 100 square ft are preferable. The Green Book recommends that sides be at least 12 ft long (15 ft preferred) after rounding corners; however, this dimension can be overly restrictive since it implies an island area of at least 80 ft. Cut-through crosswalks are preferred over ramps and should follow ADA guidance for sidewalks. The raised crossing; prominent signs; and alignment of the turn lane with good visibility of crossing pedestrians/cyclists and a non-tangential turn at the end all help promote low-speed turns and yielding compliance. The shared-use path approach angle affords ideal visibility, and the island size makes it a suitable waiting area. Where channelized right-turn lanes are not signalized, their geometry should help promote motorist yielding to crossing pedestrians and bicycles. Helpful elements include a small turning radius; a narrow, channelized roadway; raised crossings; and having the turn lane meet the cross street at a near right angle. If a raised crossing is used, detectable warning surfaces are required across the entire area that is level with the roadway. When bicycles use a delta island, their approach should avoid sharp or sudden turns and should enable them to see conflicting traffic without looking over their shoulder (contrast Exhibit 6-13 with Exhibit 6-15). Locating the crosswalk near the center of the island and turn is preferred for pedestrian sight distance and visibility. Guidance through landscaping or other features can help pedestrians who are visually impaired or have low vision cross at the correct location (Schroeder et al., 2017). Bibliography American Association of State Highway and Transportation Officials. (2011). A Policy on Geometric Design of Highways and Streets, 7th Edition (Green Book). Washington, DC.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 57   Bachiochi, J., & Furth, P. G. (2016). Commonwealth Avenue Bridge: Protecting Pedestrians and Bicycles with Pre signals and Delta Islands. http://www.northeastern.edu/peter.furth/comm-ave-bu-bridge-a-compact-intersection- based-on-a-copenhagen-design/ Manual on Uniform Traffic Control Devices for Streets and Highways. (2009). FHWA, U.S. DOT. http://mutcd. fhwa.dot.gov/ Potts, I. B., Harwood, D. W., Bauer, K. M., Gilmore, D. K., Hutton, J. M., Torbic, D. J., Ringert, J. F., Daleiden, A., & Barlow, J. M. (2014). NCHRP Web-Only Document 208: Design Guidance for Channelized Right-Turn Lanes. Transportation Research Board, Washington, DC. Schroeder, B., Rodegerdts, L., Jenior, P., Myers, E., Cunningham, C., Salamati, K., Searcy, S., O’Brien, S., Barlow, J., & Bentzen, B. L. (2017). NCHRP Research Report 834: Crossing Solutions at Roundabouts and Chan- nelized Turn Lanes for Pedestrians with Vision Disabilities: A Guidebook. Transportation Research Board, Washington, DC. 6.5 Leading Pedestrian Intervals 6.5.1 Basic Description 6.5.1.1 Alternative Names Pedestrian head start; partially protected crossing. 6.5.1.2 Description and Objective In an LPI, a Walk signal indication is shown to pedestrians a few seconds earlier than the start of green for the concurrent vehicular movement. The crossing is protected-only during this initial interval; for the remainder of the pedestrian phase, turning conflicts are allowed, making it a partially protected crossing. Exhibits 6-17 and 6-18 show the pedestrian and vehicular move- ments during and after the leading interval and how these movements are aligned in time (City of Boston, 2013). The main objective of an LPI is to enable pedestrians to arrive at the conflict point before the first right-turning vehicle in order to reinforce the priority that pedestrians have over turning vehicles. This improves pedestrian safety and comfort by promoting yielding compliance on the part of turning motorists and reducing pedestrian–vehicle conflicts. 6.5.1.3 Variations LPIs can also be leading bicycle intervals, serving bicycles as well as pedestrians, if local laws or enforcement policies support that function. (a) During the leading interval (b) During the rest of the pedestrian phase Source: City of Boston (2013). Exhibit 6-17. Pedestrian and vehicular movements involved in an LPI.

58 Traffic Signal Control Strategies for Pedestrians and Bicyclists 6.5.1.4 Operating Context An LPI might be appropriate where pedestrian crossings are concurrent with a parallel vehic- ular phase, where right turns (in this report, “right turns” also includes left turns from a one-way street) are permitted to conflict with crossings, and the following four criteria are met: 1. There is no exclusive right-turn lane. If there is a right-turn lane, it is more appropriate to use the delayed-turn technique (see Section 6.6), which holds turning vehicles while allow- ing through vehicles to run during the leading interval. (Where there is no right-turn lane, delayed turn can still be considered as an alternative to LPI; see Section 6.5.3.3.) 2. The intersection layout fails to give pedestrians an adequate head start in space. When the vehicular stop line is set back approximately 50 ft from the curb where pedestrians wait— often called a “protected intersection” layout (National Association of City Transportation Officials, 2019), as shown in Exhibit 6-19—pedestrians get a head start in space. Details on how to evaluate the head start in space afforded by a large vehicular setback are given at the end of this list. 3. Conflicting right-turn volume is low or moderate. When right-turn volume is high, there will be a lot of conflict between pedestrians and turning vehicles, even with an LPI. Therefore, in such cases it may be more appropriate to use a treatment that provides a fully protected cross- ing, such as concurrent-protected phasing (see Section 6.2), exclusive pedestrian phases (see Section 6.3), or channelized right turns and delta islands (see Section 6.4). If turn volume is low to moderate, LPI can be appropriate; however, delayed turn (see Section 6.6) may still be preferred—even where there is no right-turn lane—because it limits the impact on motor vehicle delay, thereby providing longer protected intervals for pedestrians. 4. There is any level of pedestrian volume. Where pedestrian volumes are low, pedestrian phases can be actuated—with the LPI occurring only when a pedestrian phase runs—thus avoiding unnecessary delays on vehicle traffic. An LPI might also be appropriate when these four criteria are met and there are T-junctions or junctions with one-way streets that have no opposing through traffic to shield pedestrians from left turns during the early part of the pedestrian phase. An LPI can be considered as a means of partial protection from left turns in these cases; however, a longer LPI may be required. In Dutch cities, LPIs tend to be very short (1 s or less), and most intersections have none because the typical Dutch intersection layout has a large stop-line setback that gives pedestrians a large head start in space (S. Linders, personal communication, August 2, 2018). To determine Source: City of Boston (2013). Exhibit 6-18. Pedestrian phase and its parallel vehicular phase during and after an LPI.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 59   the needed length of an LPI, engineers identify the conflict zone (i.e., the area where pedestrians and turning cars conflict) and measure two distances (see Exhibit 6-19): dPed = distance for a pedestrian to reach the middle of the conflict zone, which is then con- verted to a time, tPed, by dividing by a design walking speed, such as 3.5 ft/s. dRT = distance a right-turning car must cover to reach the near edge of the conflict zone; this can be converted to a time, tRT, by assuming a typical right-turning speed, such as 15 ft/s (10 mph). After converting those distances to times, the needed LPI duration is provided in Equa- tion 6-3 as tLPI: ( )= −max , 0 (6-3)t tPed tRTLPI That is, if tRT ≥ tPed, no LPI is needed because even without an LPI, pedestrians will be able to establish their priority in the crosswalk before turning cars reach the conflict zone; otherwise, the green start for vehicles is delayed by (tPed – tRT). Note: dPed = the distance a pedestrian covers to reach the middle of the conflict zone dRT = the distance a right-turning vehicle covers to reach the middle of the near edge of the conflict zone Exhibit 6-19. Protected intersection layout, in which pedestrians have a large head start in space.

60 Traffic Signal Control Strategies for Pedestrians and Bicyclists To illustrate the method, suppose dPed = 14 ft and dRT = 45 ft. Using the suggested speeds given earlier, tPed = 14/3.5 = 4.0 s, while tRT = 45/15 = 3.0 s. Therefore, the needed length of LPI is 4.0 – 3.0 = 1.0 s. 6.5.2 Applications and Expected Outcomes 6.5.2.1 National and International Use LPI has been applied in the U.S. since the 1990s and has become ever more popular due to its pedestrian safety benefits. New York City has implemented LPIs at more than 2,200 inter- sections, after a pilot study from pre-2011 installations found that LPIs led to a 13% decrease in pedestrian and cyclist injuries and a 62% decrease in pedestrians and cyclists killed or seriously injured in crashes involving turning vehicles (NYC Department of Transportation, 2016). (An interactive map of current locations in New York City with LPI can be found at http://www. vzv.nyc.) Cambridge, MA, uses LPIs at most of their intersections, and many other cities in the United States and Canada make extensive use of LPIs. In U.S. practice, LPI is not limited to high pedestrian-volume locations; for example, in Charlotte, NC, LPI is a standard treatment for arterials in suburban parts of the city (for more detail, see Section 6.6). In North America, the typical length of an LPI is 3 to 7 s. In New York City and Montreal, most LPIs last 7 s; in Cambridge and Washington, DC, they commonly last 3 s. In Dutch cities— where the typical protected intersection layout gives pedestrians a large head start in space— LPIs tend to be very short, often 1 s or less. As U.S. cities reconstruct intersections with corner bulb-outs, protected bike lanes, and other features that help create a substantial setback of the vehicular stop line relative to the curb where pedestrians wait to cross, they may also find that they can use very short LPIs to accomplish their objective. 6.5.2.2 Bicycle Use of LPIs By default, bicycles in the U.S. may not use LPIs because bicycles are supposed to follow vehicular signals, not pedestrian signals. New York City and Washington, DC, are exceptions, with local laws that allow bicycles to follow pedestrian signals (in Washington, this only applies during an LPI). New York’s law, enacted in 2019, followed a successful six-month pilot program in which 50 intersections were signed to allow bicycles to follow the pedestrian signal. In cities like Chicago and Cambridge (and New York before 2019)—where LPIs are common and there is a general understanding that cyclists will not be ticketed for going on an LPI— bicycle use of LPIs has become routine. Recognizing this practice, Cambridge’s traffic signal engineers implemented LPIs that are 7 s long (longer than the usual 3 s) on a street with pro- tected bike lanes, with the intention of protecting bicycles as well as pedestrians. In the Netherlands, bicycles use LPIs because they are controlled by bicycle signals, which are programmed to release bicycles simultaneously with pedestrians. In the U.S., FHWA interim guidance prevents bicycle signals from being used in connection with LPIs, since it forbids the use of bicycle signals where vehicular turn conflicts are permitted. As a result, formally allowing bike use of LPIs with bike signals would require either an exception from FHWA or a change in FHWA guidance regarding bicycle signals and permitted-turn conflicts. 6.5.2.3 Benefits and Impacts Benefits and impacts discussed in this section include: • Reduction in pedestrian crashes and injuries; • Improved motorist yielding, which makes crossings less stressful; and • Traffic capacity decrease, delay increase, and possible cycle-length increase.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 61   A comparison of 26 intersections in New York State found that LPIs reduced pedestrian crashes with turning vehicles by 28%; the reduction was 64% after adjusting for crash severity (King, 2000). The Highway Safety Manual assigns a CMF of 0.413 for applying LPI based on a study of 10 intersections in State College, PA, where a 59% reduction in pedestrian–vehicle crashes was found (Fayish & Gross, 2010). Other studies have found that LPIs reduce the per- centage of compromised pedestrian crossings (Hubbard et al., 2007); reduce the frequency of pedestrians yielding to turning vehicles (Van Houten et al., 2000); reduce the number of vehicles turning in front of pedestrians in a crosswalk (San Francisco Municipal Transportation Agency Pedestrian Program & University of California Traffic Safety Center, 2008); and reduce the number of observed pedestrian–motor vehicle conflicts (Van Houten et al., 2000). When the timing of the minor-street phase is governed by pedestrian needs, adding an LPI has almost no impact on traffic operations; it can be viewed as simply shifting some of the minor street’s unused time from the end of the phase to the start of the phase. New York City has had an aggressive program of applying LPIs to signalized crossings matching this context (D. Nguyen, personal communication, August 16, 2019). However, where the concurrent vehicular phase’s timing is dominated by vehicular needs—as is typically the case for a crosswalk parallel to a major street—adding an LPI can have a significant impact on traffic capacity, or it can require a longer signal cycle. The shorter the LPI, the smaller the impact is, which results in using short LPIs for some jurisdictions. A way to accomplish the objective of an LPI without the negative capacity or cycle-length impact is to alter intersection geometry so that pedestrians get enough of a head start in space that they need either no LPI or a very short LPI, as described earlier. A study of Boston’s Melnea Cass Boulevard corridor found that with 6 s LPIs, the cycle length would have to be 110 s in order to have enough capacity. However, by rebuilding the corners using a protected intersection layout, pedestrians would have enough of a head start in space that an LPI would not be needed, and the needed cycle length would then be only 80 s. The capacity and cycle-length impacts of LPIs are greatest where, as in this Boston example, cross streets have high traffic volumes and therefore cannot easily give up green time for an LPI. When a minor street’s volume is low, using pedestrian overlaps with LPIs (see Section 6.7) can make it possible to create an LPI for the major street without any adverse capacity impact. 6.5.3 Considerations 6.5.3.1 Accessibility Considerations LPIs should include APS to notify visually impaired pedestrians of the LPI because otherwise visually impaired pedestrians—who may rely on the sound of starting traffic to know when the green begins—may miss an LPI and begin to cross with the vehicular movement, when motor- ists are less likely to yield to them (City of Boston, 2013; Alexandria Department of Transporta- tion and Environmental Services, 2015). LPI is especially suitable at locations with a large population of older adults or schoolchildren who tend to walk slower (Saneinejad & Lo, 2015; Staplin et al., 2001) and thus need more time to establish themselves in the crossing. 6.5.3.2 Guidance The MUTCD cites LPIs as an option; it speaks of considering LPI when there are “high pedes- trian volumes and high conflicting turning vehicle volumes,” conditions repeated by several other publications (City of Boston, 2013; Van Houten et al., 2000; King, 2000; Staplin et al., 2001). However, as cities like Charlotte have found, pedestrian volume is less applicable since

62 Traffic Signal Control Strategies for Pedestrians and Bicyclists the cost of the treatment is nominal, and for pedestrian phases that are actuated, there will be an impact to traffic only when the pedestrian phase is called. Furthermore, as explained earlier, LPI is appropriate where the conflicting turning volume is moderate or low but not where it is high. With high turning volumes, full protection from conflicting turns is preferred to partial protec- tion. The right-turn volume threshold for preferring fully protected versus partially protected crossings is 200 vehicles per hour for New York City (D. Nguyen, personal communication, August 16, 2019) and 250 vehicles per hour for Boston (City of Boston, 2013). In Montreal, that threshold is 200 vehicles per hour where the crossing length is 20 m (67 ft) or more. For shorter crossings, the threshold increases to 500 vehicles per hour where the crossing length is 8 m (27 ft) or less (City of Montreal, 2017). The MUTCD (2009) recommends a minimum LPI length of 3 s. This is a reasonable minimum for traditional U.S. intersections, considering the objective of giving pedestrians a head start. However, when intersections are configured with a large stop-line setback—as in the protected intersection layout—LPIs as short as 0.5 s can be appropriate to supplement the head start in space that pedestrians have. From Dutch experience, there does not appear to be any unintended negative consequences of using short LPIs. Among techniques that provide partially protected crossings, New York City prefers delayed turn over LPI if an exclusive turn lane can be provided. Montreal prefers delayed turn even when turn lanes cannot be provided and uses LPI only when there is a moderately high turning volume and no turn lane. 6.5.3.3 Relationships to Relevant Treatments Delayed turn (see Section 6.6) is a partial protection treatment, just like LPI. Both treatments protect pedestrians from conflicting turns during an initial part of the crossing phase; however, delayed turn allows through traffic to move during the initial interval, and therefore it has less impact on traffic operations (Furth & Saeidi Razavi, 2019). As a result, delayed turn typically provides a longer protected interval for pedestrians. NTOR (see Section 6.8) should be applied to both streets whose right turns cross a treated crosswalk. In other words, for an LPI parallel to the north–south street, NTOR should apply to both the north–south street and the east–west street, since both streets face a red signal during the LPI and right-turning drivers from both streets would otherwise go. Hubbard et al. showed that without NTOR, LPI may lose its intended benefits (Hubbard et al., 2008). Where right turn on red (RTOR) is important for capacity or delay, Section 6.8 discusses how to apply NTOR only during certain parts of a signal cycle. The desire for short cycle lengths (see Section 7.1) can be in conflict with LPIs because LPIs introduce lost time into the cycle that can force a cycle length to be longer. This conflict can be mitigated using pedestrian overlaps (see Section 6.7) and by altering corner geometry to give pedestrians a greater head start in space, thereby reducing the need for a head start in time. 6.5.3.4 Other Considerations LPIs are geared mainly toward helping pedestrians establish their priority over permitted right turns during the early part of the crossing phase. Protection against permitted left turns is typically not a consideration because during the early part of the crossing phase, permitted left turns are typically blocked by opposing through traffic. However, at T-junctions and junctions with one-way streets, there is no through traffic in one direction, therefore LPIs can provide the same support against permitted left turns. At T-junctions, the needed LPI may be considerably longer, since the conflict zone with an auto approaching from one’s rear is in the far half of the intersection. Geometric treatments that promote lower speeds and yielding by left-turning

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 63   motorists—including median islands, raised centerlines, and in-street Yield to Pedestrians signs—can also be helpful in this regard. 6.5.4 Implementation Support 6.5.4.1 Equipment Needs and Features Nearly all controllers manufactured after 2005 are programmed to allow LPIs, so no special equipment is needed. 6.5.4.2 Phasing and Timing Where LPIs are used in connection with an actuated pedestrian phase, the LPI will come up only when the pedestrian phase comes up. Where left turns are permitted during the through phase, an LPI that holds traffic in one direction must be matched with a concurrent LPI that holds traffic in the opposite direction as well. Otherwise, left-turning vehicles from the direction that is not held will pose a conflict during the LPI. With LPIs, it is better for protected left-turn phases to be lagging rather than leading (City of Boston, 2013; Department of Transportation and Environmental Services, 2015; Saneinejad & Lo, 2015). This is because if the protected left-turn interval is leading, the LPI will begin as the left-turn signal changes from protected to permitted, and left-turning drivers will be tempted to continue moving because opposing through traffic continues to stand still to serve the LPI. 6.5.4.3 Signage and Striping Not applicable for this treatment. 6.5.4.4 Geometric Elements Intersection layout determines how much of a head start pedestrians get in space, and there- fore it can affect how long an LPI (i.e., head start in time) is needed. A large stop-line setback, a curb bulb-out, and a short crossing distance from the curb to the middle of the road all lead to a shorter required LPI. Bibliography City of Boston. (2013). Boston Complete Streets Design Guidelines. City of Montreal. (2017). Feux pour Piétons à Décompte Numerique. Department of Civil and Coastal Engineering. (2008). Miami-Dade Pedestrian Safety Project: Phase II Final Implementation Report and Executive Summary. University of Florida. Department of Transportation and Environmental Services. (2015). Alexandria Complete Streets Design Guide- lines. City of Alexandria. Fayish, A. C., & Gross, F. (2010). Safety Effectiveness of Leading Pedestrian Intervals Evaluated by a Before–After Study with Comparison Groups. Transportation Research Record: Journal of the Transportation Research Board, 2198(1), 15–22. Furth, P. G., & Saeidi Razavi, R. (2019). Leading Through Intervals versus Leading Pedestrian Intervals: More Protection with Less Capacity Impact. Transportation Research Record: Journal of the Transportation Research Board, 2673(9), 152–164. Hubbard, S. M., Awwad, R. J., & Bullock, D. M. (2007). Assessing the Impact of Turning Vehicles on Pedestrian Level of Service at Signalized Intersections: A New Perspective. Transportation Research Record: Journal of the Transportation Research Board, 2027(1), 27–36. Hubbard, S. M., Bullock, D. M., & Thai, J. H. (2008). Trial Implementation of a Leading Pedestrian Interval: Lessons Learned. ITE Journal, 78(10), 37–41. King, M. R. (2000, December). Calming New York City Intersections. In Urban Street Symposium Conference Proceedings.

64 Traffic Signal Control Strategies for Pedestrians and Bicyclists Manual on Uniform Traffic Control Devices for Streets and Highways. (2009). FHWA, U.S. DOT. http://mutcd. fhwa.dot.gov/ National Association of City Transportation Officials. (2019). Don’t Give Up at the Intersection: Designing All Ages and Abilities Bicycle Crossings. NYC Department of Transportation. (2016, August). Don’t Cut Corners—Left Turn Pedestrian & Bicyclist Crash Study. NYC Department of Transportation. (n.d.). Bicyclists’ Use of Leading Pedestrian Intervals: Pilot Program Results. PedSafe. (2018, June 29). Pedestrian Safety Guide and Countermeasure Selection System. http://www.pedbikesafe. org/PEDSAFE/ Saneinejad, S., & Lo, J. (2015). Leading Pedestrian Interval: Assessment and Implementation Guidelines. Trans- portation Research Record: Journal of the Transportation Research Board, 2519(1), 85–94. San Francisco Municipal Transportation Agency Pedestrian Program & University of California Traffic Safety Center. (2008). San Francisco PedSafe Phase II—Final Implementation Report and Executive Summary. Staplin, L., Lococo, K., Byington, S., Harkey, D., & TransAnalytics, L.L.C. (2001). Guidelines and Recommenda- tions to Accommodate Older Drivers and Pedestrians (No. FHWA-RD-01-051; 1495/FR; NCP 3A6a-0042). Turner-Fairbank Highway Research Center. Van Houten, R., Retting, R. A., Farmer, C. M., & Van Houten, J. (2000). Field Evaluation of a Leading Pedestrian Interval Signal Phase at Three Urban Intersections. Transportation Research Record: Journal of the Transpor- tation Research Board, 1734(1), 86–92. 6.6 Delayed Turn/Leading Through Intervals 6.6.1 Basic Description 6.6.1.1 Alternative Names Leading through arrow; leading bicycle interval. 6.6.1.2 Description and Objective At the start of a vehicular through phase, pedestrians, bicycles, and through vehicles are allowed to go while turning movements (right turns, left turns, or both) are held for an interval called a leading through interval (LTI) or leading bike interval. This interval typically lasts 7 to 13 s. After the leading interval, the hold on turning movement(s) is lifted, allowing a permitted conflict with the bicycle and pedestrian crossings. Like LPI, delayed turn provides a partially protected crossing, meaning the first part of the crossing phase is conflict-free. However, it differs from LPI in that during the leading interval, only turning traffic is held whereas the through movement is permitted. This allows a longer protection interval, since only the turning vehicles are held. Exhibits 6-20 and 6-21 show the movements allowed during and after the leading interval. The objective of this treatment is summarized as follows: • To give pedestrians a partially protected crossing with less traffic capacity impact than would occur with LPI; • To give pedestrians a longer initial protected interval than would be practical with LPI; • To give bicycles a partially protected crossing, providing the first flush of waiting bicycles conflict-free passage through the intersection (note that, in general, bicycles are not legally allowed to use leading pedestrian intervals—New York City and Washington, DC, are excep- tions that allow bicycles to legally use LPIs); and • To reduce delay to bicycles, pedestrians, and turning traffic while providing partial protection at intersections where it is safe to have permitted-turn conflicts after a conflict-free initial interval. 6.6.1.3 Variations Delayed turn can be applied with or without an exclusive turn lane for the turn(s) being held. In Montreal, delayed turn is widely applied where there are no exclusive turn lanes; in New York

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 65   (a) During LTI (b) After LTI Exhibit 6-20. Pedestrian, bicycle, and vehicular movements during and after an LTI. Exhibit 6-21. Alignment of pedestrian and concurrent vehicular phases with delayed turn/LTI. City and Charlotte, it is applied only with an exclusive turn lane. Where turning vehicles share a lane with through vehicles (e.g., in Montreal), through vehicles can be blocked by turning vehicles ahead of them during the initial protected interval. Two variations of signal displays have been used to implement delayed turn. In New York and Charlotte, where turning traffic is always in an exclusive turn lane, a red turn arrow is employed during the leading interval along with a circular green, changing to FYA (see Sec- tion 6.9) during the permitted-turn period. In Montreal, only a through green arrow is displayed

66 Traffic Signal Control Strategies for Pedestrians and Bicyclists during the leading interval, and a circular green is displayed during the permitted-turn interval (see Exhibit 6-22). Whether a through green arrow alone is enough to prohibit turning move- ments will depend on state/provincial law. 6.6.1.4 Operating Context Delayed turn may be appropriate in connection with concurrent pedestrian and bicycle crossings: • Where there is a need to provide partially protected crossing for bicycles as well as pedestrians without causing high delays for vehicles and pedestrians/bicycles; and • Where there is a low or moderate turn volume, and intersection geometry affords good vis- ibility and prevents high turning speed. – If turn volume is high, turning speed is high, or there is poor visibility for the conflict between turning cars and crossing bicycles/pedestrians, a treatment offering full protection should be considered (see Section 6.2; Section 6.3; and Section 6.4). Where there is an exclusive turn lane, delayed turn is a superior alternative to LPI. Where there is no exclusive turn lane, delayed turn can be considered as an alternative to LPI if the turning volume is low, intersection capacity is a concern, or there is a desire to have a long leading protected phase in order to improve safety and comfort for far-side crossers. Where there is a high turning percentage and there is no exclusive turn lane, blockage (i.e., turning vehicles blocking through vehicles) can be so frequent that it may be more desirable to use an LPI. 6.6.2 Applications and Expected Outcomes 6.6.2.1 National and International Use In Montreal, delayed turn—known locally as leading through arrow—has been a standard treatment for more than 15 years throughout the downtown area and at other intersections with moderate or high pedestrian-traffic (Montreal Department of Transportation, 2017; J. Hamaoui, personal communication, June 5, 2018). It is used at more than 100 intersections where the leading interval is between 7 and 13 s long. Almost none of the application sites have exclusive turn lanes. During the leading intervals, both right and left turns are held. Most of Source: Peter Furth. Exhibit 6-22. Through green arrow followed by a circular green in Montreal.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 67   the intersections where it is used have only pedestrian crossings; however, it is also used where there are bicycle crossings. In New York City, delayed turn has been used since approximately 2015—usually with a 10 s leading through interval—and always in conjunction with FYA during the permitted-turn interval. New York has at least 37 intersections where this treatment has been installed (NYC Department of Transportation, 2016). It has become the city’s preferred treatment for protected bike lane crossings wherever a short turn lane can be created (usually by removing parking) and turn volume is not high. Where turn volume exceeds 200 vehicles per hour, the city prefers to use concurrent-protected crossings (see Section 6.2). Delayed turn is also used at intersections with moderately heavy pedestrian-volumes where a turn lane can be provided (D. Nguyen, per- sonal communication, August 16, 2019). New York developed the delayed-turn treatment based on its experience with concurrent- protected bicycle crossings. When creating the nation’s first parking-protected bike lanes, they were cautious about intersection safety and used fully protected bicycle crossings. However, they found that at intersections with low turn volumes, many cyclists were running the red during the turn phase due to the long waiting times without a noticeable safety impact. This resulted in implementing the delayed-turn treatment at certain intersections to lower both bicycle and vehicle delay. Initially, the turn volume threshold for switching from full protection to partial protection was 120 vehicles per hour; after a few years’ experience, that threshold was raised to 200 vehicles per hour (D. Nguyen, personal communication, August 16, 2019). In one downtown corridor, Montreal recently changed signals from delayed turn to concur- rent-protected, which is the opposite of New York’s experience. Boulevard de Maisonneuve, a one-way street through downtown with a bidirectional cycle track on the left side, had been using delayed turn since the cycle track opened in 2007. However, as bicycle traffic grew, left- turning drivers were not finding enough safe gaps and were turning unsafely, sometimes resulting in a collision. The bidirectional nature of the cycle track—with bicycles arriving in both directions and sometimes arriving late during the green period, as modulated by upstream traffic signals— made it more difficult for turning drivers to find safe crossings. The change to fully protected crossings required converting one of the boulevard’s two lanes into a left-turn lane. Charlotte uses delayed turn at about 10 intersections, and roughly another 10 are in progress (N. Conrad, personal communication, December 26, 2018; see Exhibit 6-23 for an example). Delayed turn is the basic element of a local program called LPI+, for which the city won an ITE innovation award in 2016 (North Carolina Section of the Institute of Transportation Engineers, Source: North Carolina Section of the Institute of Transportation Engineers (2016). Exhibit 6-23. An LTI in Charlotte. During the permitted-turn interval, the red arrow is replaced by an FYA. The sign reading “No Turn on Red” is blank except during the LTI.

68 Traffic Signal Control Strategies for Pedestrians and Bicyclists 2016). Unlike New York and Montreal, Charlotte has only applied delayed turn outside of its business district, typically on wide arterials where the pedestrian phase is actuated. As part of the LPI+ program, during periods of the day with low turning volumes, the LPI encompasses the entire pedestrian phase and becomes a concurrent-protected crossing (see Section 6.2). Where no right-turn lane is available, the city uses an LPI (see Section 6.5), typically lasting 3 to 5 s and up to 10 s for some crossings at schools. Kothuri et al. (2018) report that in a survey of professionals regarding bicycle signal control strategies, out of 69 respondents, about half were aware of the delayed-turn technique, but only one respondent’s city had used it. 6.6.2.2 Benefits and Impacts Benefits and impacts discussed in this section include: • Safety and comfort as they relate to the length of the protected interval; • Inclusion of bicycles in a partially protected crossing; • Safety as observed by traffic conflicts; and • Capacity and delay impacts to vehicular traffic. For pedestrians, the delayed turn improves safety and comfort by providing them with a conflict-free head start, which enables them to establish themselves in the crosswalk before turning traffic arrives at the conflict point. This leads to improved yielding behavior by motor- ists during the permitted-turn period. This is the same benefit provided by LPIs, except that delayed turn is usually considerably longer (because it has less impact on intersection capacity) and offers greater protection. While LPIs are usually only long enough for near-side crossers to establish their priority before turning vehicles arrive, delayed turn is sometimes long enough to also enable far-side crossers to reach the conflict zone before turning vehicles. For bicycles, delayed turn improves safety and comfort, enabling the first flush of bicycles to pass through an intersection conflict-free. Note that in general, LPIs cannot be used legally by cyclists. Delayed turn can be more effective than LPI in reducing bicycle–motor vehicle conflicts. In New York City, a left-side, parking-protected bike lane runs along 6th Avenue. At its inter- section with 23rd Street, the crossing treatment was changed from LPI (which some bicycles used) to delayed turn with a 7 s leading interval. The treatment also included adding a left-turn lane (by removing parking) and FYA during the permitted-turn interval. Exhibit 6-24 shows results of a before-and-after study, with reported results normalized to indicate conflicts per 1,000 bicycles. The two most important conflicts, near misses and collisions that would have happened had there been no evasive action, both show a strong decline. There was a small increase in the frequency of cyclists who had to ride around a car, such as when a vehicle inter- rupts its turn for crossing pedestrians (Kothuri et al., 2018). Leading Pedestrian Interval Delayed Turn with Added Left-Turn Lane and FYA Bicycles 1,952 1,300 Per 1,000 bicycles: Near misses 4 0 Collisions if no evasive action taken 75 35 Bicycle rode around car 32 41 Source: Kothuri et al. (2018); the project team’s analysis of that data. Exhibit 6-24. Bicycle-vehicle conflicts with LPI versus delayed turn, 6th Avenue at 23rd Street, New York.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 69   If turns are from an exclusive lane, then delayed turn should not affect the delay for any movements other than the turns that are held during the leading interval. If turns are from a shared through-turn lane, as in Montreal, then a small amount of additional delay to through traffic should be expected, increasing with the volume of turning traffic. Kothuri et al. used a simulation to measure delay impacts of a 5 s delayed turn at a Portland intersection under various demand scenarios. The affected approach had an exclusive right- turn lane. As expected, additional delay to through traffic and to cross-street traffic was either undetectable or less than half a second. Additional delay to right-turning vehicles was also less than a second because RTOR was allowed during all red periods except the delayed turn, leading to very little queuing in the right-turn lane (Kothuri et al., 2018). A simulation study of delayed turn/LTI with no exclusive turn lanes contrasts the capacity impact of LTI versus LPI (Furth & Saeidi Razavi, 2019). The results showed that while LTI’s capacity loss increases with the proportion of right turns, its impact on intersection capacity is still far lower than the capacity loss due to an LPI, especially on multilane approaches. In one example scenario—a three-lane approach in which 20% of the traffic turns right, the rightmost through lane is shared with right turns, and pedestrian demand is such that they block the crosswalk for 10 s—an LTI of 15 s has the same capacity impact as an LPI of 3 s. The far lower capacity impact of LTI means that the leading interval can be considerably longer, affording better pedestrian and bicycle protection. 6.6.3 Considerations 6.6.3.1 Accessibility Considerations At intersections with delayed turns, APS can be helpful for notifying visually impaired pedes- trians of when the Walk interval begins. Without APS, visually impaired pedestrians may be confused by the delayed start of traffic in the nearest lane, as they tend to listen to the sound of traffic parallel to the crosswalk to know about the start of a Walk interval. The through lane is farther away and can be harder to hear. 6.6.3.2 Guidance New York City’s policy makes delayed turn a preferred treatment for crossings on roads with protected bike lanes. It is only applied where they have, or can create, an exclusive turn lane. There is no minimum turn volume for applying delayed turn. The upper limit, above which they prefer fully protected crossings, is 200 vehicles per hour (during the peak hour). Where heavy bicycle-/pedestrian-crossing activity leaves limited gaps for turning vehicles, they prefer fully protected crossings (this criterion was critical for Montreal’s recent change on Boulevard de Maisonneuve, as described earlier). The pedestrian-volume threshold, below which delayed turn is preferred and above which fully protected crossings are preferred, is 700 vehicles per hour in Manhattan and 300 vehicles per hour elsewhere. If the turn bay is too short to hold the turn queue, the city uses delayed turn where a fully protected crossing would otherwise be warranted. There is no lower limit for bicycle or pedestrian volume; unless bicycle/pedestrian volume is moderately high, they will not remove parking to create a turn lane (D. Nguyen, personal com- munication, 2019). Use of a through green arrow to prohibit turns without a red turn arrow, as practiced in Montreal, is not explicitly addressed in the MUTCD but is consistent with MUTCD (2009) guidance on traffic signals. Section 4D.04, subsection A2, states: Vehicular traffic facing a GREEN ARROW signal indication, displayed alone or in combination with another signal indication, is permitted to cautiously enter the intersection only to make the movement indicated by such arrow, or such other movement as is permitted by other signal indications displayed at the same time.

70 Traffic Signal Control Strategies for Pedestrians and Bicyclists If a state’s vehicle code is consistent with this section of the MUTCD in prohibiting turns when the only signal displayed is a green arrow, then the display used in Montreal (a through green arrow without any red arrows) should be sufficient, although an effort may be warranted to promote driver understanding and compliance. The MUTCD explicitly prohibits displaying a circular red in combination with a through green arrow. 6.6.3.3 Relationships to Relevant Treatments Delayed turn/LTI is similar to LPI (see Section 6.5) in that both offer partial protection for crossings. The main differences are that delayed turn has less traffic impact, generally has longer protected leading intervals, and can serve bicycles as well as pedestrians. However, it may require an exclusive turn lane. Delayed turn can be combined with FYA (see Section 6.9) and displayed during the permitted- turn interval. NTOR (see Section 6.8) should apply during the leading interval (at a minimum). 6.6.4 Implementation Support 6.6.4.1 Equipment Needs and Features New signal heads may be needed for turn arrows or through arrows. Wherever a red turn arrow is used, it can also be configured to include an FYA for the permitted-turn phase. 6.6.4.2 Phasing and Timing New York City’s policy is that delayed turns should be 10 s long, but they can be as short as 7 s if needed for traffic capacity. Charlotte also uses 10 s durations. Montreal’s delayed turns/LTIs range from 7 to 13 s and often match the Walk interval. The city has found that with LTIs longer than 13 s, motorist frustration at being blocked by a turning vehicle (due to lack of exclusive turn lanes) leads to complaints and unsafe behavior (J. Hamaoui, personal communication, 2018). When pedestrian or bicycle phases are actuated, delayed turn is invoked only when the pedestrian or bicycle phase gets a green phase. 6.6.4.3 Signage and Striping NTOR signs may be needed to prevent turns during the LTI. Even if a red arrow is used during the initial interval, NTOR signs may still be needed where state law does not prohibit right turn on a red arrow (e.g., Oregon, Washington, Florida, and Massachusetts) or where compliance is poor. Where RTOR is desired during other parts of the signal phase, dynamic blank-out signs can be used during the leading interval only, as applied in Charlotte (see Exhibit 6-23). If no red arrow is applied during the leading interval, as in Montreal, there is no “red” period to which RTOR laws could apply. Nevertheless, drivers accustomed to turning on red may need supplemental signage to reinforce that turns are prohibited during that interval. 6.6.4.4 Geometric Elements If implemented as in New York and Charlotte, an exclusive right-turn lane is needed to pre- vent through drivers from being blocked by turning vehicles that are held during the initial protected interval. With the Montreal-style application, no turn lane is needed unless right-turn demand is substantial.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 71   Bibliography Furth, P. G., & Saeidi Razavi, R. (2019). Leading Through Intervals versus Leading Pedestrian Intervals: More Protection with Less Capacity Impact. Transportation Research Record: Journal of the Transportation Research Board, 2673(9), 152–164. Kothuri, S., Kading, A., Schrope, A., White, K., Smaglik, E., Aguilar, C., & Gil, W. (2018). Addressing Bicycle-Vehicle Conflicts with Alternate Signal Control Strategies. National Institute for Transportation and Communities. Manual on Uniform Traffic Control Devices for Streets and Highways. (2009). FHWA, U.S. DOT. http://mutcd. fhwa.dot.gov/ Montreal Department of Transportation. (2017). Feux pour Piétons à Décompte Numérique. Guide DT2001. North Carolina Section of the Institute of Transportation Engineers. (2016). 2016 Outstanding Project of Year: Charlotte Leading Pedestrian Interval Plus (LPI+). NYC Department of Transportation. (2016, August). Don’t Cut Corners: Left Turn Pedestrian & Bicyclist Crash Study. 6.7 Pedestrian Overlaps with Leading Pedestrian Intervals and Vehicular Holds 6.7.1 Basic Description 6.7.1.1 Alternative Names None. 6.7.1.2 Description and Objective During an LPI or another short interval in which all vehicular phases are held in red, pedes- trian phases in all directions can overlap. For example, in Exhibit 6-25(a), the north–south phase has an LPI; during that LPI, the east–west pedestrian phase can be extended, overlapping the LPI. In Exhibit 6-25(b), both through vehicular phases have LPIs, and all pedestrian phases are extended by overlapping them with an LPI for the perpendicular direction. The objective is either to have longer pedestrian phases or to have more efficient phasing, which can lead to shorter cycles or can allow LPIs to be introduced with less capacity impact. 6.7.1.3 Variations Pedestrian phases can also overlap during partial vehicular holds that are not LPIs, in which at least one vehicular movement has green, but other vehicular movements that could run in parallel with it are held. An example is illustrated in Exhibit 6-26, in which a left-turn and right- turn movement are running, but the vehicular movements that normally run concurrently—in this example, northbound through and southbound left—are held, allowing the two crosswalks shown to run without conflict. An informal overlap occurs when the LPI for one street is used as part of the pedestrian phase end buffer for the previous pedestrian phase. At several intersections in Cambridge, MA, pedes- trian phases are programmed to end their FDW at the end of the yellow; this allows the following (a) Overlapping one LPI (b) Overlapping two LPIs Exhibit 6-25. Pedestrian phases overlapping vehicular holds (LPIs).

72 Traffic Signal Control Strategies for Pedestrians and Bicyclists LPI to count as part of the pedestrian phase end buffer, which enables the pedestrian phases to have a longer Walk interval compared to ending FDW at the start of the yellow. This reduces pedestrian delay and creates additional crossing opportunities for slower pedestrians. While this is not programmed in the controller as an overlap, it has the same function. 6.7.1.4 Operating Context Pedestrian overlaps with LPI and other vehicular holds can be considered: 1. Wherever LPIs are used. See Exhibit 6-25(a) (LPI on one street) and Exhibit 6-25(b) (LPI on both streets). 2. Where a minor intersecting street is dominated by pedestrian timing needs while the major street is dominated by vehicular needs. In such a situation, there might be some reluctance to provide an LPI for the major street because taking time from its phase to create an LPI could have a significant capacity impact. By having the LPI overlap the intersecting street’s pedestrian phase, it may be possible to create time for the major street’s LPI without taking any time from the major street’s vehicular phase, as illustrated in an example in Section 6.7.2. 3. During low-volume periods in which both intersecting streets are dominated by pedestrian timing needs. In such a case, pedestrian overlaps can permit a shorter cycle, as illustrated in an example in Section 6.7.2. 4. Where demand for a left turn is high compared to the demand for the opposite direction’s left turn. Running a single left turn as the only vehicular movement during part of the cycle can allow two crosswalks to overlap with it, as illustrated in Exhibit 6-26. 5. Where there are multistage crossings. Introducing a short vehicular hold with pedestrian overlaps can lengthen pedestrian phases enough to create good progression for pedestrians through multiple crossing stages. 6.7.2 Applications and Expected Outcomes 6.7.2.1 National and International Use In U.S. practice, except with exclusive pedestrian phases (see Section 6.3), pedestrian phases are generally treated as “children” of a “parent” vehicular phase; and where parent phases are in conflict, their “children” are usually treated as if they are in conflict as well. But pedestrian phases are never really in conflict with one another. The recent proliferation of LPIs has created opportunities for pedestrian overlaps that did not exist before, and have rarely been taken advantage of, as with the informal overlaps used by Cambridge, MA, described earlier. Exhibit 6-26. Pedestrian movements that can overlap a left-turn phase (example of a partial vehicular hold).

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 73   In the Netherlands, where pedestrian and bicycle phases are programmed as independent phases rather than as children of a vehicular phase, pedestrian overlaps with full and partial vehicular holds occur frequently. An example of an overlap with a partial vehicular hold is in Amsterdam, at the junction of Nobelweg (considered the north–south street) and Kammerlingh Onneslaan. When the only demand for the east–west street is for eastbound left, that vehicular movement is served alone; and the non-conflicting eastbound and northbound bicycle cross- ings, along with the pedestrian crossings next to them, run concurrently (if they have a call), as in Exhibit 6-26. 6.7.2.2 Benefits and Impacts Benefits of pedestrian overlaps with LPIs and other vehicular holds include: 1. Making it possible to introduce an LPI to a major street without reducing its capacity (if the minor street is well below capacity); 2. Shortening signal cycles in low-traffic periods when pedestrian timing needs may dominate; 3. Facilitating progression through multistage crossings; and 4. Increasing pedestrian phase lengths, reducing delay, and helping slower pedestrians. 1. Making it possible to introduce an LPI to a major street. At the intersection of a minor street with a major street, it is well understood that adding an LPI to the minor street can be done with almost no impact on traffic operations. If traffic volume on the minor street is low, pedestrian overlaps with LPIs can make it possible to also add an LPI to the major street—with its attendant safety benefits—without taking time from the major street’s green phase. The fol- lowing example illustrates this idea: Major-minor intersection, with low traffic on the minor street. Consider an inter section with an 80 s cycle whose phase diagram is as shown in Exhibit 6-27(a). To serve its traffic, the major street needs a 50 s split; the minor street could serve its traffic with a 20 s split but needs 30 s to serve the pedestrian crossing. Change intervals are 4 s: • Inserting an LPI of 4 s on the minor street by starting its green later (Exhibit 6-27[b]) will have very little impact on traffic. However, since the major street cannot afford to give up 4 s of green, there is no LPI on the major street. • Because the minor street has available capacity, it can afford to give up an additional 4 s of vehicular green time and have its vehicular green end 4 s early. During this interval, while all traffic is held for 4 s, the minor street’s pedestrian phase can still run while an LPI begins on the major street (Exhibit 6-27[c]). There is no capacity impact to the major street, whose vehicular phase is unchanged. At the same time, the pedestrian phase along the major street becomes 4 s longer than it was in the original case. • Adding a second pedestrian overlap, during the minor-street LPI, further improves pedes- trian service by making the major street pedestrian phase 8 s longer than it originally was (see Exhibit 6-27[d]). Note that this final step will not be possible if the intersection includes left-turn phases. 2. Shortening signal cycles. For periods in which vehicular volumes on both intersecting streets are low enough that pedestrian timing needs dominate, letting pedestrian phases overlap with LPIs can enable a shorter cycle length with less pedestrian delay and little impact on vehicular traffic. A simulation study of an intersection in Boston found that introducing LPIs with pedes- trian overlaps in such a situation allowed a 22% reduction in cycle length, with average pedestrian delay falling from 24 s to 18 s and no perceptible change in vehicular delay (Furth et al., 2012). 3. Facilitating progression through multistage crossings. Where there are multistage cross- ings, a short vehicular-hold interval overlapped by pedestrian phases can be the key to creating good progression for pedestrians. An example is described in Section 10.1.

74 Traffic Signal Control Strategies for Pedestrians and Bicyclists 4. Lengthening pedestrian phases, reducing delay, and helping slower pedestrians. Several of the examples already provided show how overlapping pedestrian phases with LPIs can increase the length of pedestrian phases, reduce delay, and make intersections more accessible for slower pedestrians. 6.7.3 Considerations 6.7.3.1 Accessibility Considerations As described in Section 6.5 on LPIs, APS are recommended wherever LPIs are used because otherwise visually impaired pedestrians, who often take their cue from the sound of starting traffic, will not know when the pedestrian phase begins. Without an accessible signal, they would likely begin crossing when vehicular traffic is released, exposing them to greater conflict with turning traffic. 6.7.3.2 Guidance Not applicable for this treatment. 6.7.3.3 Relationships to Relevant Treatments As described earlier in this section, this treatment relates to LPI (see Section 6.5), short cycle length (see Section 7.1), pedestrian clearance settings for serving slower pedestrians (see Sec- tion 7.4), and multistage crossings (see Chapter 10). 6.7.3.4 Other Considerations Not applicable for this treatment. (a) Without LPIs (b) LPI added to minor street (c) LPI added to major street, with overlap (d) Second overlap added, lengthening the N-S Walk phase Exhibit 6-27. Introducing LPIs and pedestrian overlaps in a way that leaves the lengths of major street (north–south) vehicular phase and minor-street pedestrian phase unaffected.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 75   6.7.4 Implementation Support 6.7.4.1 Equipment Needs and Features Not applicable for this treatment. 6.7.4.2 Phasing and Timing As an example of how pedestrian phases might be programmed as overlaps, consider the example in Exhibit 6-25(b). It might have the following phase and overlap assignments: • Phase 1: east–west LPI; • Phase 2: east–west vehicular movement; • Phase 3: north–south LPI; • Phase 4: north–south vehicular movement; • Overlap A, consisting of Phases 1, 2, and 3: east–west pedestrian movements; and • Overlap B, consisting of Phases 3, 4, and 1: north–south pedestrian movements. To effectively use the second LPI in each of the overlaps, the through phases (Phases 2 and 4) must either be pretimed or coordinated in such a way that the ending time is known in advance. Overlaps like this are only possible where a through movement immediately follows another through movement. Therefore, it is not possible to have two such overlaps where an intersection has left-turn phases. 6.7.4.3 Signage and Striping Not applicable for this treatment. 6.7.4.4 Geometric Elements Not applicable for this treatment. Bibliography Furth, P. G., Muller, T. H. J., Salomons, M., Bertulis, T. A., & Koonce, P. J. (2012). Barrier-Free Ring Structures and Pedestrian Overlaps in Signalized Intersection Control. Transportation Research Record: Journal of the Transportation Research Board, 2311(1), 132–141. Furth, P. G., Yue, D. W., & Santos, M. A. (2019). Multi-Stage Pedestrian Crossings and Two-Stage Bicycle Turns: Delay Estimation and Signal Timing Techniques for Limiting Pedestrian and Bicycle Delay. Journal of Trans- portation Technologies, 9.04, 489. 6.8 No Turn on Red 6.8.1 Basic Description 6.8.1.1 Alternative Names None. 6.8.1.2 Description and Objective NTOR refers to a restriction on right turns during red intervals that are otherwise allowed. This section is limited to consideration of NTOR to enhance pedestrian and bicycle safety. NTOR addresses not only collisions with crossing pedestrians and bicycles, but also the inconvenience and hazard that occurs when right-turning drivers block the crosswalk while checking for a sufficient gap in traffic to finish their turn. Drivers often wait with their gaze fixed to their left, which may keep them from noticing a pedestrian or bicycle approaching from their

76 Traffic Signal Control Strategies for Pedestrians and Bicyclists right—even when the crossing person is directly in front of the vehicle—creating a high-risk situation, especially for pedestrians or bicycles with a low profile such as children and people in wheelchairs. 6.8.1.3 Variations Some states allow left turns on red from a one-way street to another one-way street. Similar guidance applies for these movements as well. NTOR can be implemented with time restrictions (e.g., 7 a.m. to 7 p.m.) or “when pedestrians are present.” It can also be applied during certain phases of a signal cycle by using dynamic blank-out signs. 6.8.1.4 Operating Context Some contexts in which NTOR may be appropriate to protect pedestrians and bicycles include: • Crossings with a moderate or high pedestrian-/bicycle-volume or with a significant volume of vulnerable crossers (e.g., children or older adults); • Where the crosswalk location is such that drivers turning right block the crosswalk while waiting for a gap; and • Crossings used by bicycles approaching from the right side (e.g., from a two-way path). NTOR is also necessary in conjunction with LPI (see Section 6.5), delayed turn (see Sec- tion 6.6), exclusive pedestrian/bicycle phases (see Section 6.3), and concurrent-protected cross- ings (see Section 6.2). These treatments all aim to hold right-turning vehicles during certain phases, so NTOR should be in effect during those phases. 6.8.2 Applications and Expected Outcomes 6.8.2.1 National and International Use This treatment is widespread throughout the United States. New York City and Montreal are the only places in the U.S. and Canada that prohibit RTOR by default. Everywhere else, the restriction must be signed. In some cities—including Washington, DC; Seattle; and Alexandria, VA—adding NTOR restrictions is part of their Vision Zero plans. Several agencies—including those in Ithaca, NY; Charlotte; and Portland—use dynamic blank- out signs to apply NTOR during selected phases in the signal cycle. Charlotte uses dynamic NTOR signs to hold right-turning traffic during pedestrian-crossing phases, which are generally pushbutton actuated. Wherever it applies the delayed-turn technique (see Section 6.6), when there is a pedestrian call, the through phase begins with an interval lasting about 10 s in which, in addition to a green ball, a red right-turn arrow and a dynamic NTOR sign are displayed. For the remainder of the through phase, the dynamic sign reads “Yield to Pedestrians” and the red turn arrow is replaced with an FYA (Thomas et al., 2016). 6.8.2.2 Benefits and Impacts Several studies have shown that permitting RTOR increases crashes with pedestrians and bicycles. Preusser et al. (1982) found that RTOR increases right-turning crashes with pedestrians by 43% to 107% and with bicycles by 72% to 123%, resulting in a CMF ranging from 1.43 to 2.08. A CMF of 1.69 is given in the Highway Safety Manual for bicycle and pedestrian crashes when RTOR is changed from prohibited to permitted. However, applying NTOR restrictions has not produced evidence of a strong safety effect. According to Harkey et al. (2006), NTOR signs are estimated to reduce crashes by about 3%, with a CMF of 0.97.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 77   A review of Fatality Analysis Reporting System data showed that in the 10-year period between 1982 and 1992, less than 1% of all national traffic fatalities involved a right-turning vehicle at an intersection that permits RTOR; however, bicyclists or pedestrians were involved in more than half of those fatal crashes (NHTSA, 1994). California conducted a comparable study of its crash data and found similar results, suggesting that its current policy of selectively restricting RTOR was better than a blanket ban at all intersections (Fleck & Yee, 2002). A 2002 study in Arlington, VA, found that NTOR restrictions conditioned on time of day are far more effective than those conditioned on pedestrian presence (Retting et al., 2002). At five previously unrestricted intersections treated with signs reading, “No Turn on Red, 7 a.m.–7 p.m., Mon–Fri,” there was a large decrease in RTOR, a large increase in vehicles stop- ping before turning, and a large decrease in pedestrians yielding to vehicles turning right on red. At five other previously unrestricted intersections with fluorescent yellow-green reflective signs reading “No Turn on Red—When Pedestrians are Present,” observations only from when a pedestrian was present found only a small decline in RTOR vehicles, little change in the fraction of RTOR vehicles stopping before they turned, and little change in the number of pedestrians yielding to RTOR vehicles. A recent study conducted by Lin et al. (2016) found that driver com- pliance was the highest with signs readings “No Turn on Red” (70%) compared to “Right on Red Arrow after Stop” (67%), “Turning Vehicles Yield to Pedestrians” (67%), and “Stop Here on Red” (55%). In one study, dynamic NTOR blank-out signs used only during school crossing periods or other critical times were found to be only slightly more effective than static NTOR signs (Zegeer & Cynecki, 1986). However, a study of different treatments at a Miami, FL, intersection found dynamic NTOR signs were considerably more effective than static signs (Pecheux et al., 2009). Violations were high with two different static signs—“No Turn on Red” and “No Turn on Red When Pedestrians in Crosswalk,”—but fell significantly when dynamic signage was used. Prohibiting RTOR will increase delay for right-turn vehicles. Because right turns typically have extra capacity, it is less likely for RTOR restrictions to significantly impact traffic capacity. Where right-turn volumes are so great that prohibiting RTOR may substantially affect intersec- tion operations, adjusting the signal timing may address the issue, since vehicles turning right on red can only do so when there is a gap in the conflicting traffic, indicating potential to take some green time from the cross street. 6.8.3 Considerations 6.8.3.1 Accessibility Considerations Allowing RTOR makes it harder for visually impaired pedestrians to identify the surge of traffic at the onset of the vehicular green phase on the street parallel to the crossing direction, therefore it increases the need for APS. Because visually impaired travelers, in the absence of APS, wait to hear a vehicle traveling straight across the intersection to determine if the signal has changed, they are frequently delayed in initiating crossing where vehicles can turn right on red (Barlow et al., 2003). 6.8.3.2 Guidance The MUTCD (2009) indicates six conditions for when an NTOR sign should be considered. The conditions that include non-motorized considerations include: • Geometrics or operational intersection characteristics that might result in unexpected conflicts; • An exclusive pedestrian phase or LPI; Note: The ongoing NCHRP Project 03-136 is currently evaluating the performance of RTOR operations at signalized intersections. The project aims to develop methods and tools that consider all modes and inform planning and operations decisions for practitioners. Learn more by visiting the project webpage (https:// apps.trb.org/cmsfeed/ TRBNetProjectDisplay.asp? ProjectID=4549).

78 Traffic Signal Control Strategies for Pedestrians and Bicyclists • An unacceptable number of pedestrian conflicts with RTOR maneuvers, especially if they involve children, older pedestrians, or persons with disabilities; and • More than three RTOR crashes reported in a 12-month period for that particular approach. The MUTCD also provides guidance for the design of various types of NTOR regulatory signs. Detailed discussion is provided in Section 6.8.4. 6.8.3.3 Relationships to Relevant Treatments NTOR should be used in combination with the following treatments, either as a dynamic restriction or as a restriction of RTOR only during phases in which the treatment is active: • Concurrent-protected crossings (see Section 6.2); • LPI (see Section 6.5); • Delayed turn/LTIs (see Section 6.6); and • Exclusive pedestrian/bicycle phases (see Section 6.3). 6.8.3.4 Other Considerations If RTOR is prohibited, more vehicles will turn right during a green phase, when they could conflict with pedestrians making a concurrent crossing. This concern can be mitigated by using LPI or delayed turn. 6.8.4 Implementation Support 6.8.4.1 Equipment Needs and Features NTOR could be implemented with static and dynamic blank-out signs. There are several variations of NTOR signs allowed by the MUTCD, as shown in Exhibit 6-28. NTOR signs with a red ball (R10-11) have been shown to be more effective than standard black-and-white signs. While the MUTCD (2009) indicates that RTOR is intended for use while a circular red ball is displayed but not while a red turn arrow is displayed, state laws are not uniformly consistent with this distinction. Even where they are, many motorists often do not understand the law that applies in their state. To avoid ambiguity and improve compliance, it may be advisable to pro- vide signage for NTOR in conjunction with red arrows as well as with circular red indications. Dynamic blank-out signs can be activated during programmed phases and powered from the controller cabinet. Considerations for mounting signs near the signal head include visibility as well as weight and wind loads. 6.8.4.2 Phasing and Timing Not applicable for this treatment. 6.8.4.3 Signage and Striping Exhibit 6-28. NTOR signs (R10-11 Series) in MUTCD (2009).

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 79   6.8.4.4 Geometrics This treatment can be applied with and without exclusive right-turn lanes. NTOR is an effective countermeasure when (1) the skew angle of the intersecting roadways makes it difficult for drivers to see traffic approaching from their left or (2) geometrics or opera- tional characteristics of the intersection result in unexpected conflicts, such as between a right- turning vehicle and a left-turning vehicle entering the same departure leg. Bibliography AAA Mid-Atlantic. (2019, February 5). Seeing Red: District Moves to Ban Right Turns on Red at Nearly 100 Inter- sections in the Nation’s Capital. https://dccommutetimes.com/news_releases/article/seeing-red-district- moves-to-ban-right-turns-on-red-at-nearly-100-intersect Barlow, J. M., Bentzen, B. L., & Tabor, L. S. (2003). NCHRP Research Results Digest 278: Accessible Pedestrian Signals: Synthesis and Guide to Best Practice. Transportation Research Board, Washington, DC. City of Alexandria. (2019, June 11). New No Turn on Red Restrictions to be Installed as Part of Vision Zero Initia- tive. https://www.alexandriava.gov/tes/info/default.aspx?id=107669 City of Seattle. (2017, May). Vision Zero 2017 Progress Report. https://www.seattle.gov/Documents/Departments/ beSuperSafe/VZ_2017_Progress_Report.pdf Fleck, J. L., & Yee, B. M. (2002). Safety Evaluation of Right Turn on Red. ITE Journal, 72(6), 46–48. Harkey, D. L., Tsai, S., Thomas, L., & Hunter, W. W. (2006). Pedestrian and Bicycle Crash Analysis Tool (PBCAT): Version 2.0 Application Manual (No. FHWA-HRT-06-089). Office of Safety Research and Development, FHWA. Highway Safety Manual. (2010). AASHTO, Washington, DC. Lin, P. S., Wang, Z., Guo, R., & Kourtellis, A. (2016, May). A Pilot Study on Interactions between Drivers and Pedestrian Features at Signalized Intersections—Using the SHRP2 Naturalistic Driving Study Data. In Pro- ceedings of the 11th Asia Pacific Transportation Development Conference and 29th ICTPA Annual Conference, 70–77. Manual on Uniform Traffic Control Devices for Streets and Highways. (2009). FHWA, U.S. DOT. http://mutcd. fhwa.dot.gov/ NHTSA. (1994). Impact of Right Turn on Red (DOT HS 808 200). Washington, DC. Pecheux, K., Bauer, J., & McLeod, P. (2009). Pedestrian Safety and ITS-Based Countermeasures Program for Reducing Pedestrian Fatalities, Injury Conflicts, and Other Surrogate Measures Draft Zone/Area-Wide Evalu- ation Technical Memorandum. Preusser, D. F., Leaf, W. A., DeBartolo, K. B., Blomberg, R. D., & Levy, M. M. (1982). The Effect of Right-Turn- on-Red on Pedestrian and Bicyclist Accidents. Journal of Safety Research, 13(2), 45–55. Retting, R. A., Nitzburg, M. S., Farmer, C. M., & Knoblauch, R. L. (2002). Field Evaluation of Two Methods for Restricting Right Turn on Red to Promote Pedestrian Safety. ITE Journal, 72(1), 32–36. Thomas, L., Thirsk, N. J., & Zegeer, C. V. (2016). NCHRP Synthesis 498: Application of Pedestrian Crossing Treat- ments for Streets and Highways. Transportation Research Board, Washington, DC. Zegeer, C. V., & Cynecki, M. J. (1986). Evaluation of Countermeasures Related to RTOR Accidents that Involve Pedestrians. Transportation Research Record: Journal of the Transportation Research Board, 1059, 24–34. 6.9 Flashing Pedestrian and Bicycle Crossing Warnings 6.9.1 Basic Description 6.9.1.1 Alternative Names FYA; flashing right-hook warning. 6.9.1.2 Description and Objective A flashing pedestrian/bicycle crossing warning is a flashing signal that warns turning motorists of a possible conflict with crossing pedestrians or bicycles and reminds them of their obligation to yield. Some types of flashing warnings include images of a pedestrian or a crosswalk.

80 Traffic Signal Control Strategies for Pedestrians and Bicyclists 6.9.1.3 Variations FYA is the only commonly used flashing warning in the U.S. FYA is widely used to warn left- turning motorists of potential conflicts with oncoming traffic. FYA can also be used with right turns—and with left turns from a one-way street—to warn of conflicts with crossing pedestrians and bicycles. Several cities—including New York, Charlotte, Portland, and Boston—use FYA in connection with right turns and with left turns from a one-way street. Exhibit 6-29 shows an application in New York City. A custom flashing sign (see Exhibit 6-30) called a flashing right-hook warning is used at one intersection approach in Portland that has a history of crashes between bicycles and right- turning vehicles. Because of the downhill, bicycles tend to approach at high speed, making it harder for drivers to detect an approaching bicycle. Source: NYC Department of Transportation (2016). Exhibit 6-29. FYA in New York warning of a permitted conflict with adjacent bicycle and pedestrian crossings. Source: Andersen (2015). Exhibit 6-30. A flashing sign warning turning vehicles to yield to bikes, Portland.

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 81   A flashing right-hook warning is also used at one intersection approach in Amsterdam. Its graphics are based on a commonly used static pedestrian/bicycle crossing warning sign. Exhibit 6-31 shows the static sign and the flashing sign, which both display the image of a cross- walk with a large, red exclamation sign. This video shows the flashing sign in operation: https:// www.youtube.com/watch?v=FJQEiNNSegw (Furth, 2019a). Amsterdam uses a flashing yellow light at a few intersections to warn left-turning motorists of crossing pedestrians and bicycles. The light, substantially larger than a yellow ball used at traffic signals, is oriented diagonally, positioned to catch the attention of a driver whose vehicle has begun to turn left. Exhibit 6-32 shows the flashing warning light. This video shows the flashing sign in operation: https://youtu.be/A1JmWMIjsg0 (Furth, 2019b). This kind of signal is not (a) Static warning sign (b) Flashing warning sign (c) Close-up of flashing sign Exhibit 6-31. Static and flashing pedestrian-bicycle crossing warnings in the Netherlands. “Let op” means “Watch out.” Exhibit 6-32. Flashing yellow warning light, Amsterdam.

82 Trafc Signal Control Strategies for Pedestrians and Bicyclists Source: Boot et al. (2015). Exhibit 6-33. Flashing pedestrian indicator proposed by Florida State University researchers. currently allowed in the U.S. because of confusion from displaying a ashing yellow ball (even if it is outside a signal head) at the same time as a solid green. University researchers in the U.S. have proposed dierent forms of ashing pedestrian/ bicycle crossing warnings and tested them in a laboratory setting. e warning developed and tested at Florida State University (Boot et al., 2015) is shown in Exhibit 6-33—a yellow arrow alternates every 0.5 s with a white walking man. e concept is to alternate a common symbol of warning with a symbol indicating the object of the warning. While it is questionable whether this conguration would be granted approval by FHWA due to confusion over the ashing white man (it normally indicates the Walk interval, but the ashing warning would run throughout the pedestrian phase, including the FDW interval), the concept may have promise. At the University of Wisconsin, research by Noyce et al. (2017) found that the ashing sign shown in Exhibit 6-34 was the most promising design among alternatives tested to alert drivers who were making permitted le turns to potential conicts with pedestrians and bicycles. It is a modied version of the MUTCD’s R10-15 sign, which ashes only during permissive le-turn periods. 6.9.1.4 Operating Context Flashing crossing warnings are used where permitted right turns or le turns conict with crossing bicycles or pedestrians. ey can be used at intersections both with and without exclu- sive turn lanes. 6.9.2 Applications and Expected Outcomes 6.9.2.1 National and International Use Of the various ashing crossing warnings shown, only FYA is used routinely. Many U.S. cities use this treatment. In New York, it is a standard treatment where permitted outside turns (right turns or le turns from a one-way street) cross a bike lane, and it is used at more than 100 inter- section approaches. FYA is also frequently used where there are only conicting pedestrians. e other warning signs shown earlier are used at only one or a few intersection approaches or have been used only in laboratory experiments. 6.9.2.2 Benets and Impacts e ashing warning sign used in Portland (see Exhibit 6-30) was found to reduce right-turn conicts from 18 to 6 per 24 hours of daytime operation. e reduction in serious conicts Exhibit 6-34. Flashing pedestrian-crossing warning design by Noyce et al. (2017).

Treatments that Reduce or Eliminate Conflicts with Turning Traffic 83   (i.e., conflicts that required substantial braking or course adjustments by one or both vehicles) was even more pronounced (Andersen, 2015). Florida State University’s laboratory study of its proposed warning device found that at inter- sections with the device, drivers showed greater caution and searched more thoroughly for pedestrians. Using a questionnaire, they also found that it increased participants’ recognition of their obligation to yield when turning from 68% (without the flashing sign) to 87% (with the sign) in one test, and from 79% to 94% in another. At the same time, there was a large increase in stopping for pedestrians when pedestrians were not present, an undesirable outcome. In the University of Wisconsin’s study of its proposed warning device, 74% of respondents fully understood the intended message of the signal shown in Exhibit 6-34 (Noyce et al., 2018). Three other alternatives tested resulted in worse understanding. At Oregon State University, a laboratory study using a driving simulator found that driver- yielding behavior improved significantly when FYA was used instead of a steady green ball at locations with high volumes of permissive right turns from exclusive right-turn lanes (Hurwitz et al., 2018; Jashami et al., 2019). They found that the probability of responding correctly was 0.95 when a driver encountered an FYA versus 0.74 when the signal display was circular green. FYA is used extensively across the country in connection with permitted left turns, where turning motorists are obligated to yield to opposing traffic as well as crossing pedestrians and bicycles. While studies have shown that FYA reduces crashes with opposing motor vehicles, Van Houten et al. (2012) found that the safety benefit to pedestrians is insignificant. FYA is part of the delayed-turn treatment described in this guidebook, for which researchers found a significant reduction in conflict rate (see Section 6.6). However, it is not possible to assess how much of that reduction stems from using an FYA as opposed to other features of the treatment—that is, a leading, protected interval and the addition of a turn lane. 6.9.3 Considerations 6.9.3.1 Accessibility Considerations Not applicable for this treatment. 6.9.3.2 Relationships to Relevant Treatments FYA is combined with delayed turn (see Section 6.6) in New York City and Charlotte and is sometimes combined with LPI (see Section 6.5) as well. 6.9.4 Implementation Support 6.9.4.1 Equipment Needs and Features FYAs are a standard signal display that may have to be added. Portland’s and Amsterdam’s flashing warning signs are custom signs, powered from the controller cabinet and controlled there so as to be active only during the relevant phase. 6.9.4.2 Phasing and Timing Flashing crossing warnings, including FYA, should be active at any time in the cycle in which pedestrians or bicycles are expected to be in conflict with a turning movement. Where a conflict is only expected during a pedestrian phase, it is advisable to continue the flashing sign from when the pedestrian phase goes to solid Don’t Walk until the end of the concurrent vehicle phase, since pedestrians may still be clearing during the early part of the solid Don’t Walk interval.

84 Traffic Signal Control Strategies for Pedestrians and Bicyclists 6.9.4.3 Signage and Striping Not applicable for this treatment. 6.9.4.4 Geometric Elements Not applicable for this treatment. Bibliography Andersen, M. (2015, April 2). Right-Hook Risk Drops with Flashing Yield to Bikes Sign on NE Couch. https:// bikeportland.org/2015/04/02/right-hook-risk-drops-flashing-yield-bikes-sign-ne-couch-136458 Boot, W., Charness, N., Roque, N., Barajas, K., Dirghalli, J., & Mitchum, A. (2015). The Flashing Right Turn Signal with Pedestrian Indication: Human Factors Studies to Understand the Potential of a New Signal to Increase Awareness of and Attention to Crossing Pedestrians (No. BDV30-977-13). Florida Department of Transportation. Furth, P. G. (2019a). Flashing Right Hook Warning in Amsterdam [Video]. YouTube. https://www.youtube.com/ watch?v=FJQEiNNSegw Furth, P. G. (2019b). Flashing Yellow Crossing Warning, Amsterdam [Video]. YouTube. https://youtu.be/A1Jm WMIjsg0 Hurwitz, D., Monsere, C., Kothuri, S., Jashami, H., Buker, K., & Kading, A. (2018). Improved Safety and Effi- ciency of Protected/Permitted Right Turns in Oregon (FHWA-OR-RD-18-14). Portland, Oregon Department of Transportation. Jashami, H., Hurwitz, D. S., Monsere, C., & Kothuri, S. (2019). Evaluation of Driver Comprehension and Visual Attention of the Flashing Yellow Arrow Display for Permissive Right Turns. Transportation Research Record: Journal of the Transportation Research Board, 2673(8), 397–407. Noyce, D. A., Chitturi, M. V., Alsghan, I., Santiago, K. R., & Bill, A. R. (2017). Evaluating Countermeasures to Improve Pedestrian and Bicycle Safety. Safer-Sim. NYC Department of Transportation. (2016, August). Don’t Cut Corners: Left Turn Pedestrian & Bicyclist Crash Study. Van Houten, R., LaPlante, J., & Gustafson, T. (2012). Evaluating Pedestrian Safety Improvements (Report RC-1585). Michigan Department of Transportation. 98–101.

Next: Chapter 7 - Treatments that Reduce Pedestrian and Bicycle Delays »
Traffic Signal Control Strategies for Pedestrians and Bicyclists Get This Book
×
 Traffic Signal Control Strategies for Pedestrians and Bicyclists
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

In the United States, traffic signal timing is traditionally developed to minimize motor vehicle delay at signalized intersections, with minimal attention paid to the needs of pedestrians and bicyclists. The unintended consequence is often diminished safety and mobility for pedestrians and bicyclists.

The TRB National Cooperative Highway Research Program's NCHRP Research Report 969: Traffic Signal Control Strategies for Pedestrians and Bicyclists is a guidebook that provides tools, performance measures, and policy information to help agencies design and operate signalized intersections in a way that improves safety and service for pedestrians and bicyclists while still meeting the needs of motorized road users.

Supplemental to the report are presentations of preliminary findings, strategies, and summary overview.

READ FREE ONLINE

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!