National Academies Press: OpenBook
« Previous: 4. Exploration of Right-Turn-On-Red Models
Page 65
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 65
Page 66
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 66
Page 67
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 67
Page 68
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 68
Page 69
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 69
Page 70
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 70
Page 71
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 71
Page 72
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 72
Page 73
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 73
Page 74
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 74
Page 75
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 75
Page 76
Suggested Citation:"5. Data Collection and Analysis." National Academies of Sciences, Engineering, and Medicine. 2023. Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance. Washington, DC: The National Academies Press. doi: 10.17226/27264.
×
Page 76

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.

5. DATA COLLECTION AND ANALYSIS 5.1 Data Collection Methods This chapter describes the methods of data collection employed for this study. The objective was to assemble a dataset with broad geographic coverage of the United States that would include a representative sample of RTOR operations at a collection of locations with a variety of characteristics. For purposes of comparison, a previous NCHRP study on stop-controlled intersection LOS (Kyte et al. 1996) captured video recordings between one and two hours in length to analyze a total of 79 two-way stop-controlled intersections and 49 all-way stop- controlled intersections. Those intersections were located in metropolitan areas within five geographic regions of the United States. One of the institutions comprising the team for this research is a private company that performs vehicle counts for agencies, primarily using video collected on site. The original research plan for this project had intended to collect new video from various sites across the United States by utilizing locations where orders for traffic counts were expected to be received in 2020, which would have likely included several hundred locations. These plans were made infeasible by the COVID-19 pandemic. In March of 2020, it had become apparent that pandemic conditions would last longer than the intended time frame for data collection. A period of lockdowns followed by a long period of working from home led to a substantial reduction in traffic volumes across the United States. The vast majority of agencies were not interested in collecting volumes during these exceptional conditions. Therefore, the team was deprived of the opportunity to collect new video during traffic count activities, and reduced traffic volumes meant that even if new video observations were collected expressly for this study, there was a chance it might not capture relevant conditions. An alternative to new data collection was to use archived video from past field data collection activities from 2016 through 2019. About 1,500 study locations were available for each year. These locations were not limited to signalized intersections. Some of the archived video was recorded at unsignalized intersections and road sections on unsignalized facilities. To scan through thousands of videos in a timely manner, a systematic process was used to drill down to the most relevant data. Screen captures were obtained from each study location and visually filtered to identify which videos contained recordings of signalized intersections. After this initial selection, the intersection locations were investigated further to confirm that RTOR was a possible maneuver at the intersection. Next, locations with poor video quality were excluded from the dataset. Another important consideration was that the signal heads needed to be in view so that the signal state for the right-turn movement could be determined. This led to the exclusion of numerous videos. Finally, the videos were examined to ensure that there was a substantial right-turn volume. Videos in which the right-turn volumes appeared to be very low were excluded. At the end of this process of filtering and data selection, videos from 219 intersections were obtained from the archived videos. 65

Some resources remained available for new video data collection. The research team collected new video at 33 intersections, while video from a further eight intersections was obtained with the help of agencies that volunteered to record video using existing infrastructure, such as pan- tilt-zoom (PTZ) cameras installed at the intersections. It was challenging to find locations having cameras installed at a vantage point that would include all of the movements of interest along with a view of the signal heads. In addition, some of the locations where new video data were obtained did not end up having sufficient right-turn volume to be usable for the study. Wherever possible, these opportunities for new video data collection were used to increase the representation of underrepresented types of locations, such as locations with dual right-turn lanes. Altogether, video data from a total of 260 different locations were obtained. Figure 26 shows the distribution of these locations across the United States. The dataset includes locations in every region of the United States, although certain regions are more heavily represented because of the geographic distribution of agencies that requested traffic counts for which the archived video was originally recorded. 6 (MD) 3 (DC) 1 Figure 26. Heatmap showing distribution of data collection locations. 5.2 Data Analysis Because of the variety of cameras used across the various locations, the quality of the video collected for this research varied considerably among different locations. Much of the archived video utilized a fisheye lens to get a wider view of the intersection in the frame; these tended to 66

have a lower resolution than newer video. Agency video from PTZ cameras or other existing field infrastructure was of similar quality in some cases, but in other cases the quality was slightly better. New video data were collected by cameras with better resolution. Figure 27 shows example views from different video sources. (a) Video collected using higher resolution camera. (b) Video with fisheye lens used in traffic count studies. (c) Video captured by existing field infrastructure. Figure 27. Example video views. The original research plan had intended to employ machine learning-assisted analysis techniques to reduce the video and thereby obtain a larger quantity of video for a lower amount of effort. Unfortunately, much of the archived video did not have sufficient resolution to permit the use of automated techniques for counting individual movements and determining the signal states. Instead, a high-level analysis was undertaken in which the time periods with the heaviest volumes (determined by a preliminary analysis with machine learning assessment) were selected 67

for manual observations. Videos ranged in duration from 2 to 12 or more hours. The busiest 2 hours were selected in each case for the manual observations. Human observers recorded 5-minute counts of vehicles and pedestrians departing an intersection. Separate counts of vehicles and pedestrians were recorded for green and red intervals. Durations of red time were recorded for all of the intervals from each video, and cycle lengths were noted for a few intervals. A sample of raw data from one hour is shown in Table 13. Subscripts R and G in Table 13 represent counts during red and green durations, respectively. Volumes V1, V2, V3, P1, P2, and LT correspond to the arrows in Figure 14 on Page 45. Table 13. Sample data from Gaines School Road and Cedar Shoals Drive, Athens, Georgia t R C RTOR V1R V2R LTR V3R P1R P2R RTOG V1G V2G LTG V3G P1G P2G 0 151 120 20 0 9 45 0 0 0 25 0 2 0 0 0 0 5 123 7 1 3 25 0 0 0 31 0 4 0 0 0 0 10 133 11 0 6 29 0 0 0 34 0 8 0 0 0 0 15 139 120 13 1 11 30 0 0 0 36 0 4 0 0 0 0 20 163 20 0 9 51 0 0 2 12 0 4 0 0 0 0 25 124 9 0 6 24 0 0 0 33 0 6 0 0 0 1 30 176 120 13 1 12 42 0 0 0 27 0 7 0 0 0 0 35 132 14 0 8 35 0 0 0 41 0 3 0 0 0 0 40 162 13 0 14 40 0 0 1 37 0 2 0 0 0 0 45 132 120 11 0 10 31 0 0 0 43 0 5 0 0 0 0 50 177 31 1 11 37 0 0 0 26 0 1 0 0 0 3 55 162 15 0 14 22 0 0 0 41 0 3 0 0 0 0 Note: t = start of 5-minute interval; R = total red duration in seconds; C = average cycle length in seconds; RTOG = right-turn-on-green. The rest of the notations are similar to those used in Figure 14. 5.3 Characteristics of the Included Intersections A variety of site characteristics for the intersections were compiled based on observations from online aerial and street view images for each location (in addition to the evidence contained in the camera views). For each location, the following attributes were determined: • Whether the intersection is an interchange ramp terminal • The presence of bus stops (near side or far side, and on the subject or crossing approach) • Presence of other transit facilities at the intersection • Presence of parallel parking (on the subject or crossing approach) • Presence of parallel or conflicting pedestrian crosswalks • Presence of bicycle lanes (on the subject or crossing approach) • Type of RTOR treatment (typical with stop on red, channelized with yield sign, channelized with free movement) • Type of RTOR geometry (single exclusive right-turn lane, shared through and right-turn lane, dual exclusive right-turn lane) • Number of receiving lanes • Number of conflicting through lanes 68

• Presence of shared through and right-turn lanes on the conflicting approach • Presence of shared through and left-turn on the opposing left-turn movement • Configuration of the shadowed left-turn lane (dedicated lane, shared through and left turn, prohibited turn) • Presence of right-turn arrow (typically indicating a right-turn overlap) • Presence of signage (RTOR permitted after stop, RTOR when no pedestrians present, yield to pedestrians in crosswalk) • Configuration of opposing left turn (protected, permitted, or protected-permitted) • Whether the intersection is in a central business district or similar area (such as a college campus) • Presence of median on crossing street • Subject and crossing approach speed limits • Presence of significant intersection skew A selection of site attributes is summarized in Figure 28. These represent data for the 260 locations for which all of the site attributes have been confirmed. 69

Interchange Yes No 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Bus Stops Yes No 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Lane Configuration Single Dual Shared 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Channelization Yes No 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Conflicting Thru Lanes One More than One 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Opposing Left Turn Lanes One More than One 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Receiving Lanes One More than One 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Shadowed Left Turn Yes No 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Right Turn Overlap Yes No 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Opposing Left Turn Protected Protected-Permitted Permitted 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Parallel Conflicting Pedestrian Crosswalk Both None 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Parallel Conflicting Bicycle Lane Both None 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Figure 28. Summary of site characteristics. 70

Out of 260 locations, there were 165 with a single exclusive right-turn lane, 27 with dual right- turn lanes, and 68 with a shared through and right-turn lane on the subject approach. The total numbers of data points for these lane configurations were 5,391, 721, and 2,016, respectively. This amounts to a total of approximately 677 hours of video. Figure 29 shows histograms of RTOR flow rate for each lane configuration. The RTOR volumes for all three configurations are skewed to the right (i.e., positively skewed). The mean RTOR volume is greater than the median for all three lane configurations. As would be expected, the mean RTOR volume for the dual right-turn lanes was the highest, while the mean RTOR volume for the shared right-turn lanes was the lowest. There was greater variability of RTOR count for dual right-turn lanes compared to the exclusive and shared right-turn lanes. 71

1600 (a) Single 1400 Number of Records 1200 1000 800 600 400 200 0 0 200 400 600 800 1000 1200 1400 RTOR Flow Rate 120 (b) Dual 100 Number of Records 80 60 40 20 0 0 200 400 600 800 1000 1200 1400 RTOR Flow Rate 1000 900 (c) Shared 800 Number of Records 700 600 500 400 300 200 100 0 0 200 400 600 800 1000 1200 1400 RTOR Flow Rate Figure 29. Distributions of RTOR count for different lane configurations. Some of the major characteristics of the dataset were explored. The RTOR flow rate for any lane configuration is expected to have an upward trend with the total right-turn flow rate. Therefore, an initial step was to check whether this trend was observed in our dataset. Figure 30 shows scatterplots of RTOR flow rate and total right-turn flow rate for different lane configurations. In all cases, there is an upward trend in the RTOR flow rate, as expected. 72

1400 (a) Single 1200 RTOR Flow Rate (veh/h) 1000 800 600 400 200 0 0 100 200 300 400 500 600 700 800 Total Right Turn Flow Rate (veh/h) 2500 (b) Dual RTOR Flow Rate (veh/h) 2000 1500 1000 500 0 0 200 400 600 800 1000 1200 1400 Total Right Turn Flow Rate (veh/h) 1200 (c) Shared 1000 RTOR Flow Rate (veh/h) 800 600 400 200 0 0 100 200 300 400 500 600 700 Total Right Turn Flow Rate (veh/h) Figure 30. RTOR flow rate versus total right-turn flow rate for different lane configurations. As discussed in the previous chapter, RTOR may occur during three different intervals within a cycle at a typical intersection. One of these intervals is the green for the conflicting through movement, and another is the green for the opposing left-turn movement. During these intervals, right-turning vehicles on the subject approach will be able to execute the turn if sufficient gaps are available in in the conflicting and opposing traffic. Therefore, the RTOR flow rate should 73

decrease with increases in the conflicting through and opposing left-turn volumes. The data show a negative trend, as shown in the scatterplots in Figure 31 and Figure 32. 800 700 Total RTOR Flow Rate (veh/h/ln) 600 500 400 300 200 100 0 0 500 1000 1500 2000 2500 Conflicting Through Flow Rate (veh/h/ln) Figure 31. Total RTOR flow rate versus conflicting through flow rate. 800 700 Total RTOR Flow Rate (veh/h/ln) 600 500 400 300 200 100 0 0 100 200 300 400 500 600 700 Conflicting Opposing Left Turn Flow Rate (veh/h/ln) Figure 32. Total RTOR flow rate versus conflicting opposing left-turn flow rate. The third RTOR interval serves the shadowed left-turn movement on the intersecting street. Locations featuring a right-turn overlap will typically have a green indication during this 74

interval. However, in the absence of an overlap, the RTOR maneuver can be made more or less without any conflicting traffic, although drivers are expected to come to a stop before executing the turn, so the saturation flow rate will be lower than during green. The dataset shows a positive trend between these variables, as shown in Figure 33. 800 700 Total RTOR Flow Rate (veh/h/ln) 600 500 400 300 200 100 0 0 50 100 150 200 250 300 350 400 450 500 Shadowed Left Turn Flow Rate (veh/h/ln) Figure 33. Total RTOR flow rate versus shadowed left-turn flow rate. Finally, the scatterplot of RTOR flow rate versus total pedestrian flow rate in Figure 34 shows a negative trend, since RTOR vehicles need to wait for pedestrians in the parallel and conflicting pedestrian crosswalks before executing the turn. 75

800 700 Total RTOR Flow Rate (veh/h/ln) 600 500 400 300 200 100 0 0 50 100 150 200 250 300 350 400 450 Conflicting Pedestrian Volume (ped/h) Figure 34. Total RTOR flow rate versus total conflicting pedestrian flow rate. In general, the trends in these four sets of charts are not very strong because many additional factors contribute to the RTOR flow rate, such as the total right-turn volume. Instead of focusing on apparent correlations in the point cloud, it makes more sense to examine the upper bound of the RTOR flow rate, which suggests the trend in capacity, as a function of the other volumes. Finally, each of these diagrams shows the total RTOR volume as opposed to the component of the RTOR volume during the constituent intervals; this additional detail could not be consistently obtained during the video analysis because not all signal heads were in view at each intersection. This information would be needed to break the total RTOR volume down into separate intervals. 5.4 Conclusion This chapter presented the outcomes of the data collection and analysis activities undertaken during this study. Altogether, data from 260 locations across 25 states and the District of Columbia were assembled to support development of models of RTOR volume and capacity. Descriptive statistics were developed to understand the distributions of different quantitative variables. Trends between the RTOR flow rate and variables that may influence it were observed to check whether they make sense. The next chapter presents the development of RTOR volume and capacity estimation models. 76

Next: 6. Model Development, Calibration, and Validation »
Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance Get This Book
×
 Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

The determination of the level-of-service (LOS) at signalized intersections is an important activity for decision-making in the allocation of resources for managing public roads, estimating the impact of new developments, and designing signal timing plans. There is a need to develop models of right-turn-on-red (RTOR) volume to permit users of the Highway Capacity Manual methodology to estimate the RTOR rather than rely on collection of field data, which often does not include RTOR as a separate quantity.

NCHRP Web-Only Document 368: Right-Turn-on-Red Operation at Signalized Intersections with Single and Dual Right-Turn Lanes: Evaluating Performance, from TRB's National Cooperative Highway Research Program, addresses these needs through the development of models for RTOR volume prediction and the development of improved guidance for whether to allow RTOR.

The document is supplemental to NCHRP Research Report 1068: Right-Turn-on-Red Site Considerations and Capacity Analysis: Practitioner's Guide.

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!