Application of Pedestrian Crossing Treatments for Streets and Highways (2016)

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.

61 CHAPTER FIVE EXAMPLES OF GUIDANCE TOOLS AND ORIGINAL CASE EXAMPLES ON PROVISION OF SAFER PEDESTRIAN CROSSINGS EXAMPLES OF STATE AND LOCAL JURISDICTIONS CROSSING TREATMENT GUIDES A number of local jurisdictions as well as a few states have developed guidance to help practitioners standardize their approaches and considerations for determining appropriate actions in regard to pedestrian crossings. Boulder, Colorado, crossing treatment guidance was updated in 2011 to reflect current research including evalu- ations and experiences with treatments in that city (City of Boulder Transportation Division 2011). A number of juris- dictions besides Boulder have found value in the guide, as mentioned by survey respondents, and it is summarized here using information directly from the guide. A flowchart and guidance for North Carolina DOT was rolled out in 2015, following pilot trainings and feedback from across the state. A summary of that guide is also pro- vided following the description of the Boulder guide. Again, the information presented comes directly from the report and flowchart. Finally, because many jurisdictions are interested in how to assess locations for PHBs, Tucson, Arizona, which first implemented PHBs as an experimental device in the 1990s, provided the comprehensive criteria and ranking weights it uses to assess locations for PHBs. Boulder, Colorado—Pedestrian Crossing Treatment Installation Guidelines (2011) See https:/www-static.bouldercolorado.gov/docs/pedestrian- crossing-treamtment-installation-guidelines-1-201307011719. pdf. In November 2011, the city of Boulder updated its crite- ria, procedures, and policies for the installation of pedestrian crossing treatments. The guide is intended to help assist the city with long-established goals to provide safe and efficient pedestrian facilities that will help its goal to reduce auto- dependency. The decision to travel as a pedestrian is in part subject to the pedestrian’s ability and perceived ability to safely and efficiently cross roadways along the travel route. The new guidelines combine data collected for the previ- ous 1996 City of Boulder Pedestrian Crossing Treatment Warrants document and new data collected for the 2011 updated document. The new guide proposes updated pedes- trian crossing treatment criteria and procedures, which are selected using a flowchart and decision matrix approach, and identifies specific treatments that are applicable for the range of pedestrian and vehicle volumes, types, speeds, and road- way geometry. Since 1996 new studies and research have become available, which has allowed the city to update its approach toward pedes- trian crossing treatments and safety. The city has used results from Zegeer et al. (2005) Safety Effects of Marked Versus Unmarked Crosswalks at Uncontrolled Locations: Executive Summary and Recommended Guidelines, as well as 14 years of data collected at 40 locations with a variety of treatments to outline a new process for selecting and evaluating treatments. Data collection has taken place on certain commonly applied treatments that include enhanced signage, pedestrian-actu- ated flashing signs, and raised crossings on right-turn bypass islands. These have been installed at two-lane and multilane intersection and midblock crossings. Results have revealed that while the devices have typically increased driver yield- ing, some treatments have also led to higher vehicle-to-vehicle and vehicle-to-pedestrian crashes at multilane high-volume locations. The combination of these results and recent national and international research was used to create a “consistent pro- cedure for considering the installation of crossing treatments where needed on a case-by-case basis in the City of Boulder. Implementation of crossing treatments will require funds that could potentially have been spent on other transportation sys- tem improvements, and, therefore must be considered carefully in the funding allocation process” (City of Boulder Transporta- tion Division 2011, iv). The new evaluation procedure involves a four-step process: Step 1—Identification and Description of Crossing Loca- tion: The pedestrian crossing location site characteristics should be identified. These include the major and minor street of the crossing, if the crossing connects both ends of a multiuse path, the posted speed along the major street, and any existing traffic controls or crossing treatments, mark- ings, lighting, or relevant physical features. If the cross- ing connects both ends of a multiuse path, then minimum pedestrian volume requirements are not required to be met to apply prescribed treatments.

62 Step 2—Physical Data Collection: This involves deter- mining the existing roadway configuration, number of lanes, presence of a median, and presence of and distance to the nearest marked or protected crossing, and measuring the stopping sight distance on all vehicular approaches to the crossing. If the sight distance is less than eight times the posted speed limit, then removal of obstructions or lowering the speed limit or both would be considered. Step 3—Traffic Data Collection and Operational Observa- tions: This involves the collection of pedestrian volumes dur- ing peak hours, differentiated by important locational features like proximity to schools, and separate counts for pedestrians and bicyclists, elderly and young, and disabled pedestrians. Hourly and average daily traffic counts of vehicles also should be accomplished. Finally, specific traffic flow and operational observations are to be recorded, which include significant queuing, vehicle types, and directionally specific queues. Step 4—Apply Data to Flowchart and Decision Matrix: Formal fill-in-the-blank worksheets are provided that allow the recording of the data in Steps 1–3. Once these data are gathered, they can be used to apply a flowchart process and decision matrix to determine the proper type and implemen- tation of specific crossing treatments. The flowchart first asks if the crossing is at a controlled or uncontrolled location. Then, through a series of yes-or-no answers to specific questions regarding site data collected in Steps 1–3, multiple treatments or actions are determined (Figure 26). Possible treatments or actions at uncontrolled crossing sites include the following: 1. No action recommended 2. Consider unmarked pedestrian crossing facilitation 3. Direct pedestrians to nearest marked or protected crossing 4. Consider HAWK, traffic signal, or grade-separated crossing 5. Remove sight distance obstructions or lower speed limit, or both 6. Go to criteria for crossing treatments and uncon- trolled locations decision matrix. Possible treatments or actions at controlled crossing sites include the following: 1. No action recommended 2. Install marked crosswalk 3. Install marked crosswalk with school crossing sign on mast arm 4. Install marked crosswalk with advanced pedestrian signs FIGURE 26 Flowchart for pedestrian crossing treatments—uncontrolled or controlled crossing location. Source: Pedestrian Crossing Treatment Installation Guidelines, City of Boulder Transportation Division (2011).

63 5. Consider neck downs, median refuge, or additional signage to increase driver awareness 6. Install marked crosswalk with school pedestrian crossing sign and down arrow at crosswalk, plus advance signs. A decision matrix is to be used for identifying treatments at uncontrolled locations. It uses roadway configurations, presence of refuge islands, number of lanes, vehicle volume per day, and posted speed to identify treatments types. See Table 1 in Boulder’s guide, available at https:/www-static. bouldercolorado.gov/docs/pedestrian-crossing-treamtment- installation-guidelines-1-201307011719.pdf. Several overall issues are to be analyzed in addition to the flowchart and decision matrix process. These include the stipulation that treatments would not be included at locations with less than 1,500 vehicles per day with the exception of school crossing locations where peak hour traffic exceeds 10% of ADT. There is also a minimum threshold of roughly 20 pedestrians per hour, where pedestrian crossing treat- ment should be applied. This has exceptions for the young and elderly and school crossings at 18, 15, and 10 per hour. Each possible pedestrian treatment is defined in terms of what it must include each time it is installed. Pedestrian sig- nals, HAWKs, and RRFBs also have further selection criteria detailed by linear graphs with detailed thresholds based on total pedestrians per hour, speed limits above 35 mph and at or below 35mph, and total vehicles per hour (see Figure 27). Finally, supplemental policies provide overall guidance. These include details on adequate crosswalk lighting, avoid- ing overuse of crossing treatments, multiuse path crossing signage and crosswalk marking requirements, textured and colored pavement treatment details, accessible crosswalk requirements, use of raised crosswalks at right-turn bypass islands, and the removal of treatments. This process has been developed toward the city’s over- all goals of providing safe and effective pedestrian cross- ing treatments and continuing to evaluate and improve its existing network, and doing so within the bounds of the financial and physical constraints for implementation. The data collection, flowchart, decision matrix, and supple- mental policies are based on recent state-of-the-practice research as well as local data collection results over the past 14 years at 40 sites. The guidance helps simplify the process for choosing the most effective treatments depend- ing on the relevant site-specific characteristics of individ- ual locations. North Carolina Department of Transportation (NCDOT)— Pedestrian Crossing Guidance and Flow Chart, Schroeder et al. (2015) See https:/connect.ncdot.gov/resources/safety/Teppl/Pages/ Teppl-Topic.aspx?Topic_List=P37. In April 2015, NCDOT released a Pedestrian Crossing Guidance Flow Chart designed to provide a systematic pro- cess for engineers and planners to use in the evaluation of which crossing locations warrant facilities, and what kinds of enhancements would be provided in a Complete Streets context. In particular, this research focused on guidelines for NCDOT to evaluate the feasibility of including crosswalks or pedestrian signalization or both at signalized intersec- tions and marked crosswalks on the approaches of uncon- trolled intersections (Schroeder et al. 2015). The proposed flowchart process is to be used in conjunc- tion with field visits to verify site characteristics and gather data. Engineering judgment is to be applied with flowchart use. Crossing locations within school zones are beyond the scope. A 95-page guide supplements the flowchart with sup- porting information, warrants, images, and references. The flowchart is to be used when citizen or municipal requests are made, during development of pedestrian or gre- enway plans, upon identification of a pedestrian crash hot spot, as part of systematic reviews of existing crossing loca- tions, and as a component within an established operations and maintenance assessment process. FIGURE 27 Pedestrian signal, RRFB, PHB-HAWK selection line graphs for speeds of 35 mph and less (top), and speeds of more than 35 mph (bottom). Source: Pedestrian Crossing Treatment Installation Guidelines, City of Boulder Transportation Division (2011).

64 Important data variables to be gathered at crossing loca- tions are traffic volume, speed limit, operating speed, quan- tity and type of pedestrian activity, pedestrian crash history, roadside features and conditions, area factors, and signing or other traffic control device options. The flowchart is broken up into four steps (Figure 28). Step 1—Document Existing Characteristics/Signalized Crossing Assessment: This step focuses on whether to provide pedestrian signals at signalized crossings. The first action in Step 1 is to gather relevant data, which includes the previously listed important variables. In addition, other data variables to be included are sight distance restrictions, driver yield- ing rates, pedestrian compliance, crash history, heavy truck traffic, lighting, proximity to transit stops, special pedestrian populations (e.g., children and elderly), future projected traf- fic or pedestrian volumes or both, and future projected land use changes, growth, or development patterns (Figure 29). FIGURE 28 Overview of NCDOT four-step assessment process. Source: Schroeder et al. (2015). FIGURE 29 NCDOT Flow Chart Step 1. Source: Schroeder et al. (2015).

65 In the process of moving through the flowchart in Step 1, possible outcomes could be no action required, to install pedestrian signal heads, to consider installing signal heads, or to move to Step 2 of the flowchart. When estimating pedes- trian volume, choosing an appropriate low-volume threshold is required based on site characteristics. Issues include if the crossing location is not near high pedestrian trip generators, does not connect complementary land uses, or has less than 25 pedestrians per hour during peak hours or less than 100 pedestrians per day, or if there are significant populations of children and elderly pedestrians. Step 2—Unsignalized/Midblock Crossing Assessment: If the location is an unsignalized intersection or midblock loca- tion, Step 2 begins with determining the number of lanes, the posted or operating speed, vehicle traffic volumes, and pedestrian volumes. Possible outcomes include no action required, consider marking the crosswalk, and moving to Step 3, which addresses treatments in addition to or as an alternative to marked crosswalks (Figure 30). Number of lanes does not take into account lane width or on-street parking and bike lanes that increase crossing distance. Two- way center turn lanes are to be counted as one lane when determining lane count. Most sites can use the posted speed limits; however, when there is concern that 85th percentile operating speeds may be near or exceed speed thresholds for treatment selection, a speed study is to be conducted to determine the actual 85th percentile operating speed. Pres- ence of raised medians is also to be considered. Multilane undivided roads or painted medians are considered the same as no median. It is important that engineering judgment be used to determine which marking pattern is most appropri- ate, but midblock crosswalks are always to be marked with high-visibility patterns. School crossings as well as locations where pedestrians may not be expected by drivers can also use high-visibility markings. Step 3—Additional/Alternative Treatments Assessment: If it is determined in Step 2 that the location requires treat- ments in addition to or in lieu of marked crosswalks, then Step 3 begins with determining if the posted or 85th percen- tile operating speed is less than or equal to 35 mph or greater than 35 mph (Figure 31). Pedestrian volumes, MUTCD signal warrants, estimated pedestrian delay, and motorist compliance are also important variables. Potential outcomes include considering geometric improvements, signal instal- lations, marked crosswalks, supplemental treatments, or moving to Step 4, which assesses the possibility for PHBs. Possible geometric improvements require further engi- neering studies to determine which crossing geometries are to be implemented. These may include installation of median refuge islands, curb extensions, traffic-calming devices, treatments to minimize crossing distance and make the crossing perpendicular to traffic, and treatments that enhance visibility of and by the pedestrian by removing obstacles to pedestrian and driver lines of sight. Depending on if MUTCD Warrants 4 or 5 are met, traffic signal instal- lation may be appropriate but not required. Engineering studies should be used to determine the appropriate traffic signal pedestrian enhancement. Total pedestrian delay for the approach is determined using Equation 18-21 of the 2000 Highway Capacity Manual and multiplying that by the peak- hour pedestrian volume. Motorist compliance is considered high if driver yielding to pedestrians rates in the area are high. Speed limit, motorist compliance, and total pedestrian FIGURE 30 NCDOT Flow Chart Step 2. Source: Schroeder et al. (2015).

66 delay is used in a decision matrix to decide whether only marking the crosswalk would be considered; if supplemental warning signs, actuated beacons, or RRFBs are to be con- sidered in addition to markings; or moving to Step 4 of the flowchart should be initiated. Step 4—Pedestrian Hybrid Beacon Assessment: The final step is used to determine if a PHB should be considered. It begins with checking whether the posted or 85th percentile operating speed is less than or equal to 35 mph, or greater than 35 mph (Figure 32). The crosswalk length, vehicle volume, and pedestrian volume are then used to determine whether to consider a PHB or another signal installation, or if no action is required. Line graphs based on pedestrians per hour and vehicles per hour, with crossing length as the curve, are used to plot the specific location. If the plotted point falls above the appropriate curve for crosswalk length, then installing a PHB is to be considered. This assumes that a traffic signal does not meet MUTCD Warrants 4 or 5, or a traffic signal was rejected using an engineering study. If the plotted point falls below the curve for crosswalk length, then supplemental warning signs, markings, or RRFBs would be considered. Tucson, Arizona—Criteria and Ranking for PHBs Many jurisdictions are interested in how to assess locations for pedestrian hybrid beacons. Tucson, which first imple- mented PHBs as an experimental device in the 1990s, pro- vided the updated comprehensive criteria and weights it uses to assess and rank locations. Tucson receives many requests to install these devices as the public has become familiar with them. The city developed a comprehensive list of crite- ria and points for potential PHB installation, which is repro- duced as Figure 33. Using the criteria listed, the city had a 2014 ranked list of 92 sites for potential PHBs, with points from a high of 85 (ranked first) to 20 points. ORIGINAL CASE EXAMPLES With the assistance of the jurisdictions, six original case examples were developed for this synthesis. In addition, a few case examples are included that were originally devel- oped for other guidance resources, including the Safe Routes to School Guide and PEDSAFE. The case example from Seattle describes the use of the FHWA crosswalk study guidance to assess all uncontrolled, marked crosswalks in the city to identify if and what treatments were needed to improve their safety. Short- and long-range plans for addressing the crosswalk deficiencies were developed. The District of Columbia Department of Transportation (DDOT) pedestrian program adapted the FHWA crosswalk FIGURE 31 NCDOT Flow Chart Step 3. Source: Schroeder et al. (2015).

67 study guidance by Zegeer et al. (2005) to the lower speed limits of Washington, D.C., and developed a tailored deci- sion matrix with four treatment categories for different types of roads. In addition, DDOT adapted warrant information from Tucson, Arizona, for PHBs. Eugene, Oregon, has used multiple forms of outreach to identify problems and solutions that help meet the commu- nity’s goals of environmental sustainability and an equitable street network that fits with its goal to make it easier for people to walk. Cambridge, Massachusetts, uses a design approach to calm streets and encourage safe sharing by all modes. The city evaluates traffic-calming projects through speed stud- ies and neighborhood feedback. To implement such projects, the city incorporates traffic-calming projects into planned reconstruction projects and other project types in a Com- plete Streets approach. San Francisco and New York City are both Vision Zero cities and have taken data-driven and community-based planning approaches to identify problems and the most cost-effective treatments to drive down pedestrian fatali- ties and injuries. Doubtless there are many other good examples of effec- tive practices in outreach and identification of problem locations, and application of best available knowledge to identify effective solutions to provide safer pedestrian cross- ings. Following the original case examples, a few other case examples describing innovative treatments and approaches are replicated from their original sources. These provide a few examples of how a thoughtful approach to assessing the FIGURE 32 NCDOT Flow Chart Step 4. Source: Schroeder et al. (2015).

68 FIGURE 33 Tucson, Arizona, criteria and points scale for ranking sites for potential PHB application. Source: Diahn Swartz, Tucson Transportation Department.

69 problem can lead to new designs and applications of opera- tional measures that can create safer and context-appropri- ate solutions. Seattle, Washington—Pedestrian Crosswalk Inventory and Improvement Program Adapted from “The City of Seattle, WA, USA, Crosswalk Inventory and Improvement Plan” by Hefferan and Lagerwey (2004), with supplemental information provided by Toole Design Group and Seattle Department of Transportation In 2001, the city of Seattle used the results of the study, Safety Effects of Marked Versus Unmarked Crosswalks at Uncon- trolled Locations by Zegeer et al. (2005) to inventory and sys- tematically evaluate approximately 850 uncontrolled marked crosswalks in its jurisdiction. Seattle used the inventory and evaluation process to identify crosswalks that did not need any changes, crosswalks that could be made safer by adding treat- ments, and a small number of crosswalks that were removed. The results of the Zegeer et al. study provided guidance on effects of marking crosswalks at uncontrolled locations. This report found conditions for which marked crosswalks appear to have no effect on the number of pedestrian crashes, but it also found that higher pedestrian volumes and ADT, and a greater number of lanes, were all associated with higher rates of pedestrian crashes, as described earlier. Using this information, Seattle city officials first devel- oped a crosswalk inventory form to be used on each uncon- trolled, marked crosswalk. The form was designed to be completed in the field by trained evaluators in 15 to 20 min- utes, and enabled them to determine whether each crosswalk complied with the Zegeer study guidelines. A matrix was developed based on • Number of travel lanes and median type, • Vehicle ADT, and • Posted speed limits. The matrix categorized the location as either a candidate for a marked crosswalk, not a good candidate unless mitigat- ing treatments were included, or may or may not be a good candidate for a marked crosswalk and may require mitigat- ing treatments based on the findings in Zegeer et al (2005). The form also provided a place to sketch the crosswalk and its site characteristics, included fill-in-the-blank or multiple choice answers to questions about important road variables, and was accompanied by digital photographs. This information allowed city records about each cross- walk’s current site conditions and important features for planning improvements to be updated. For example, if a loca- tion had no curb ramp, it was noted as a location for a future ADA ramp; if there was a drain, it was noted it may not be the best place for a curb extension. Information on pedes- trian and vehicle flows and patterns could also be noted. Finally, the form included a place for a suggested action plan to be identified and certain engineering measures to be sug- gested. These included sign and paint upgrades, curb ramps, bus stop relocation, additional overhead signs, illumination upgrades, curb bulb-outs, reduced curb radii, pedestrian refuge crossing islands, road diet, parking removal, raised intersections, suggested pedestrian half or full signal addi- tions, and crosswalk removal. Associated land use features like nearby schools, bus routes, and potential pedestrian gen- erators or attractors were also described for each crosswalk. Approximately 850 uncontrolled marked crosswalks were inventoried on both arterial and nonarterial streets, constituting all of the uncontrolled marked crosswalks in the city. More than 700 were found to be candidates for marked crosswalks, 40 were found to be possible candidates, and about 80 did not rank as candidates for marked crosswalks alone. Almost every crosswalk was determined to be a can- didate for at least one of the suggested improvements. When spatially analyzed, some crosswalks that were candidates for improvements were found to cluster in corridors. This aided in subsequent project prioritization, as crosswalks in corridors could often be addressed by one corridor treat- ment, rather than many different changes to individual crosswalks. For example, road diets could be used to reduce four-lane roads to two-lane roads, making it easier to add median islands, wider sidewalks, and other pedestrian safety improvements that would reap greater safety results. Given the previously cited inventory results, an effective Crosswalk Improvement Plan was developed by the Seattle Department of Transportation (Hefferan et al. 2002). The three primary goals of the Crosswalk Improvement Plan were to improve facilities at uncontrolled marked crosswalks to facili- tate increased driver yielding, to make changes that would help pedestrians cross the street more safely, and to minimize the number of crosswalks that would be removed. The long- term goal was to eliminate crosswalks in the “N” (usually not a good candidate for a crosswalk unless mitigating measures are taken) and “P” (may or may not be a good candidate for a marked crosswalk; might need mitigating measures) catego- ries, “through the provision of engineering solutions to improve the safety of such crosswalks” or through “provision of other satisfactory ways for pedestrians to get across the street” (Hef- feran and Lagerwey 2004, p.45). The plan divided inventoried crosswalks into three types of interventions based on cost, time of implementation, and complexity: • Those with simple measures such as curb ramps, • Those with moderately complex measures such as curb bulb-outs, and • Those with complex measures such as road diets.

70 Existing policies and guidelines, leveraging improvements as part of other larger road projects (termed “piggy-backing”), and connecting improvements to existing neighborhood plans also helped determine project prioritization. Projects with an over- all cost of generally more than $70,000 fell under the umbrella of the Capital Improvement Plan (Figure 34). Projects with the least amount of time and resources required were likely to be completed first; those with long timelines and high amounts of resource required were more likely to require a longer timeline and more coordination. Some improvements occurred even before the entire inventory was completed. Crosswalk remov- als were not completed without provision of other engineering treatments nearby, such as addition of traffic signals, crossing islands, curb revisions, corridor traffic calming, and other mea- sures. The plan produced a matrix that outlined projects for the next 5 years. In 2002 alone, 58 new curb ramps were installed, 800 school signs were replaced, eight locations received curb extensions, two roads diets were completed, four new traffic signals were added, and 12 new crosswalks were marked. Seattle’s Crosswalk Inventory and Improvement Program has the advantage of being systematic and based on solid data on safety effects. Before the inventory, crosswalk improve- ment locations came to light primarily through community complaints or crash data analysis projects, which may miss crosswalks that do not meet current guidelines but also do not have a history of collisions. The inventory approach is a systemic (proactive) method other jurisdictions can emulate to identify and address safety improvement opportunities at uncontrolled marked crosswalks. Washington, D.C.—Uncontrolled Crosswalk Engineering Treatments Selection Information compiled from the survey responses with supplemental information provided by George Branyan, DDOT The Washington, D.C., District Department of Trans- portation utilizes numerous federal and local guides and studies in consideration of decisions on pedestrian safety needs at crossing locations. Among federal comprehensive sources, the MUTCD and Safety Effects of Marked Versus Unmarked Crosswalks (Zegeer et al. 2005) are important influences for identifying locations for pedestrian crossing improvements. However, the MUTCD applies the Zegeer study results only on roadways with posted speed lim- its of 40 mph or higher. Because many D.C. streets have posted speeds of 30 mph or lower, the guidance needed to be adapted for this context. The District of Columbia Pedestrian Master Plan of 2009 provides a Pedestrian Design Guidelines Manual, which includes a decision matrix called the Uncontrolled Crosswalk Engineering Treatments Table. This selection matrix was developed to apply recommendations of the Zegeer study to uncon- trolled crossing locations with posted speed of 30 mph or less in Washington, D.C. (Figure 35). The matrix uses the roadway configuration (number of lanes and presence of median) and the vehicle volumes per day (vpd), to categorize which engineering treatment is rec- ommended. All traffic volumes below 1,500 vpd, regardless of roadway configuration, receive a parallel crosswalk and W11-2 pedestrian warning sign assembly. The remaining four columns allow engineering treatment selection based on roadway configuration at traffic volume ranges of 1,500 to 9,000 vpd, 9,000 to 12,000 vpd, 12,000 to 15,000 vpd, and greater than 15,000 vpd. Four possible treatments can be selected: • Treatment A is a high-visibility crosswalk and Side of Street Pedestrian Law Sign. • Treatment B is and in-street pedestrian stop sign with advanced stop line/bars, which are suggested to be used for all multilane crossings. • Treatment C includes an activated pedestrian device, such as an RRFB, flashing beacon, or in-roadway lights. • Treatment D involves adding a signal (PHB or full traf- fic signal) or grade separation. The matrix is used by FIGURE 34 Seattle crosswalk assessment matrix adapted from FHWA study. Source: Hefferan and Lagerwey (2004).

71 some traffic engineers and technicians as guidance and not necessarily as a strict decision matrix. Locations in need of pedestrian crossing treatments were identified in the District of Columbia Pedestrian Master Plan, including eight problematic corridors, one in each ward of the city. Several of these have been the focus of extensive traffic safety studies utilizing traffic, crash, and safety data. Many uncontrolled crosswalk locations causing con- cern are also brought to light by resident complaints, sometimes through city council members. One challenge in this process is that different departments may handle complaints differently. Some engineers and technicians rely on more traditional and conventional guidance to evaluate uncontrolled crosswalk safety problems, whereas others use the crosswalk matrix in their assessment. Ide- ally, the engineer or technician will collaborate with the pedestrian coordinator to develop a plan to improve safety at a crosswalk. DDOT’s process for determining the suitability of PHBs was developed in 2014. Requests for PHBs are sent to the signals team in the Traffic Operations Administration and a preliminary evaluation screening is conducted. The screen- ing tool is similar to those developed in Tucson and Phoe- nix, Arizona, and Portland, Oregon. The screening tool considers multiple factors, including the number of lanes; ADT; distance to nearest signalized crossing; presence of bus stops and ridership; vehicle, pedestrian, and bicyclist crashes; proximity to schools and recreations centers; and senior population. The tool has a 200-point scale, and if the locational evaluation yields a score of around 150, it triggers a full warrant study under MUTCD Pedestrian Volume Sig- nal Warrant 4 standards. As the uncontrolled crosswalk engineering treatment matrix indicates, in Washington, D.C., PHBs are considered for use on uncontrolled crossings only: • Five- and six-lane roads with no median and 12,000 to 15,000 vpd; • Six-lane roads with a median and volumes of 12,000 to 15,000 vpd; or • Four- to six-lane roads with no median and volumes higher than 15,000 vpd. The District of Columbia Pedestrian Master Plan of 2009 further states, “The HAWK signal is best suited for uncon- trolled crossings of multi-lane, higher speed or volume road- ways where there is a need to provide occasional pedestrian crossings without inordinate delay to motor vehicles (i.e., school crossings, low volume neighborhood street crossings of high volume, multi-lane arterials)” (Toole Design Group 2009). DDOT also follows MUTCD pedestrian volume sig- nal warrant guidelines when initially considering a traffic control signal of any kind (Treatment D in the matrix). DDOT’s Uncontrolled Crosswalk Engineering Treat- ments Table is tailored for the lower-speed environment of Washington, D.C., and is meant as guidance to augment other standards such as those in the MUTCD and other documents. FIGURE 35 Uncontrolled Crosswalk Engineering Treatments. Source: DDOT Pedestrian Master Plan of 2009.

72 In general, identifying locations in need of pedestrian cross- ing safety treatments is an ongoing and evolving process for DDOT, but the matrix is a good example of one proactive tool DDOT uses to guide uncontrolled crossing treatment decisions criteria. Eugene, Oregon—Community-Based Identification of Needs and Solutions Information compiled from Eugene, Oregon, survey responses with supplemental information provided by Reed Dunbar, Bicycle & Pedestrian Planner, Eugene Eugene, Oregon (population about 160,000) has adopted use of a Bicycle and Pedestrian Master Plan, a series of Trans- portation Improvement Policies, NACTO design guidelines, and 8 80 Cities vision for a more environmentally sustain- able and equitable street network that fit with the city’s goal to make it easier for people to walk. The city uses a vari- ety of means to identify crossing locations in need of safety improvements and to develop treatments that meet pedes- trian needs throughout the city. Staff frequently uses crash analysis to identify problems, as well as consistently review- ing pedestrian needs within the processes for all types of transportation projects developed. Leveraging all types of projects includes using the city’s American’s with Disabilities Act (ADA) transition plan and special funding to identify and make other improvements during repaving and ramp upgrades. The city is currently in the second 5-year cycle of a voter referendum–approved pavement repair funding plan and program and updates ramps automatically when road repaving/repairs occur. The city has a methodology that prioritizes Safe Routes to School locations first, followed by residential and commercial uses. The city prioritizes meeting not just the letter, but the spirit of ADA requirements for providing an accessible network for all pedestrians when improvements are made. In coordinating with the road maintenance plans, staff looks not just at the ramps required to be updated/installed through these projects, but also at nearby locations to ensure that other ramps needed to ensure connectivity in the area are present and up to standards. The city also has a database of sidewalk facilities, and staff knows where the gaps are. At the same time, the city leverages opportunities associated with these repaving/ramp upgrades as well as other types of projects to examine where people are crossing and whether they have sufficient time at signal-controlled locations or sufficient gaps in traffic at uncontrolled locations to safely complete crossings. The city is also engaged with the community and pro- vides a complaint hotline phone number for the public to make requests, and sometimes uses online Wikimaps or web-based systems to gather input on problem locations and types. Between phone calls and Internet-based inputs, last year the city processed more than 30,000 requests for service from community members describing where they have prob- lems and the nature of the problems. Personnel receiving the calls are trained to help gather complete and consistent types of information. The information from callers is logged into a complaint database, assigned to a project manager, and dis- seminated to all public works staff in various departments, and resolutions are tracked. The city estimates that about half of pedestrian-related calls relate to uncontrolled locations where assistance is needed to get safely across the street, with the other half relating to intersection issues such as signal timing or con- flicts. The staff works hard to investigate and follow up on all complaints, although there is not always time to investigate them all. The data can also be compared with needs identi- fied through other processes. Staff also identifies locations in a more proactive fash- ion based on land uses, traffic volumes, and vulnerable populations. Planning staff has been working with poor and non-English-speaking neighborhoods to help build bridges to those who typically do not call in complaints or needs. Improving communications and trust with underserved neighborhoods remains a challenge; in the meantime, the city’s management prioritizes staff spend- ing more time to identify needs in those communities where residents do not speak out. City councilors from those districts are also typically active in ensuring a “fair share” to address the safety needs in those communities. In addition, staff performs numerous traffic speed studies, and systemwide are looking at how fast traffic is mov- ing, because speed has a strong bearing on injury poten- tial. Many arterials in the city are posted at 45 mph (and designed for those speeds), so the city is also exploring how to push down speeds to reduce fatalities and injuries in a new Vision Zero way of thinking. To home in on the most appropriate treatments, planning staff generally performs a pedestrian count to determine how many users there are. However, if there is not an existing crossing at a severe problem location (and thus low numbers of pedestrians), staff may use pedestrian generators and land use to supplement counts with estimates of demand or unmet pedestrian need. However, there is sometimes disagreement about using such data between planning and engineering departments. The city holds a public meeting, and based on the problem types, other studies—which often include traffic gaps analysis—and public input, traffic engineers will select countermeasures. Staff also considers pedestrian characteristics to help determine appropriate treatments. For example, if the location is near an area with many seniors, the solution for a multilane roadway is more likely to include a stop-control, pedestrian hybrid beacon compared with a rectangular rapid flash beacon warning device.

73 Although the city has had success using RRFBs in con- junction with pedestrian refuge islands at multilane locations, selected in part because of lower cost, it is now considering PHBs more often on multilane streets because of the recent move toward a Vision Zero framework. For the first PHB imple- mented on a multilane roadway that separated a school from a residential neighborhood, it was installed without a pedestrian refuge at a three-lane cross section, and is reportedly working quite well (Figure 36). The location also has an active transit stop that local commuters use to get to a community college far- ther on the outskirts of town. Interestingly, this location also has an aging, grade-separated crossing that is non-ADA compliant, having only stairs for access. Families with young children and strollers, wheelchair users, and others could not use the over- pass. Speed limits in the area also are transitioning from a 55 mph limit (just out of town) to a 35 mph limit, but many drivers do not lower their speeds to 35 mph, even though the four lanes of arterial highway narrows to three lanes near the school. With all of these good reasons to use a PHB, the organization and groundswell of community support in the neighborhood helped push the PHB for this location to the top of alternatives. According to the city, “all other locations generally fea- ture some attempt to narrow the crossing distance (curb extensions or islands) and the use of pedestrian warning signs or crosswalks. Budget is also a consideration.” To some extent, improvement priorities also consider geographic or socioeconomic areas equity. FIGURE 36 Crossing enhanced with pedestrian hybrid beacon in Eugene, Oregon. Source: Reed Dunbar, City of Eugene. The processes used by Eugene illustrate the complexity and dynamism of developing safer pedestrian crossings that will meet the community’s needs. Dedicated and knowledge- able staff, responsiveness to local complaints, and other pub- lic input help provide budget-sensitive solutions appropriate to the context and community. Although the community has not adopted a definite Vision Zero approach to guide decision making, it is moving in that direction with recent decisions about PHBs versus RRFBs and a focus on lowering speeds. Cambridge, Massachussets—Traffic-Calming Project Development and Evaluation Information compiled from survey responses with supplemental information provided by Juan Avendano, Bill Deignan, and Cara Seiderman, the City of Cambridge Community Development Department The city of Cambridge runs an extensive traffic-calming project evaluation program that involves formal postpro- ject speed studies and resident surveys. The main goal of the city’s traffic-calming efforts is to design streets in such a way as to “improve the quality of life in neighborhoods and allow cars to peacefully coexist with other modes of transportation” (City of Cambridge, Community Develop- ment Department, Traffic Calming Program 2011). This is accomplished by using various physical engineering countermeasures that encourage people to drive more slowly and increase the comfort and safety of walking and bicycling, and without relying on compliance with traf- fic control devices such as signs and signals, or on speed enforcement. The implementation of traffic-calming proj- ects is largely done in conjunction with streets that are being reconstructed as part of the city’s Five Year Side- walk and Street Reconstruction Plan. This plan identifies specific streets and sidewalks that are to be reconstructed each year for the next 5 years. In addition, the city con- siders locations to implement traffic-calming measures at the request of community residents and in conjunction with other reconstruction projects such as street repaving or sewer maintenance. (Figure 38 shows the locations of projects implemented over the past 20 years.) FIGURE 37 Brattle Street before and after traffic-calming project. Source: Brattle Street Traffic Calming Project Evaluation 2012, City of Cambridge website (https://www. cambridgema.gov/~/media/Files/CDD/Transportation/ TrafficCalmingProjects/Brattle%20Street%20Traffic%20 Calming%20Project%20Evaluation20120815.pdf?la=en).

74 FIGURE 38 Traffic calming over past 20 years. Source: City of Cambridge, Community Development Department. The Five Year Sidewalk and Street Reconstruction Plan is a living document that is updated regularly as conditions change, and is based around a Complete Streets concept. This concept emphasizes that streets are for users of all ages and modes, including pedestrians, bicyclists, motorists and public transportation users. The plan states that “complete streets make it easy to cross the street, walk to shops, and bicycle to work . . . make buses run on time and make it safe for people to walk to and from train stations” (Cambridge Department of Public Works 2015). The plan’s stated goals not only emphasize Complete Streets, but outline the need to maintain safe and accessible streets, to prioritize these proj- ects based on need, and to effectively communicate design and construction projects with neighborhoods while facili- tating an integrated design process with minimum disrup- tion to community life. As part of the 5-year plan’s project prioritization, issues identified by staff and the community are also taken into account to help identify potential projects. Issues identi- fied by the community are received through a formal traf- fic-calming request form that residents submit to the city’s traffic-calming project manager. The form asks for the street names and locations of concern; the time of day that the problems occur; if other residents on the street have the same concerns; and a list of problems experienced that includes speeding, street crossing difficulty, lack of courtesy to bicy- clists, cars parked too close to the corner, difficulty biking, and drivers not yielding to pedestrians (Traffic Calming Request Form n.d.). City departments then work together to prioritize the list of potential sites based on factors such as the severity of the speeding problem; the ability to coor- dinate improvements with other construction projects; and the proximity to schools, playgrounds, and other land uses important for pedestrian safety (City of Cambridge, Com- munity Development Department 2000). Once a street location is up for reconstruction as part of the Five Year Plan, officials work closely with the neighborhood residents through a series of public meetings to determine the most effective way to slow traffic and address specific pedestrian and bicyclist safety concerns. Street width, traffic volume, and traffic speeds are studied to help determine the most effective methods of calming. The most common treat- ment types include horizontal and vertical measures, such as curb extensions, raised crosswalks, raised intersections, chi- canes, and pedestrian refuge crossing islands. After a project proposal is finalized, it is implemented in conjunction with other street reconstruction projects and coordination with other important city departments. The city has had a successful traffic-calming program for nearly 20 years. The program has constructed approximately 50 raised crosswalks and intersections in that time. The pro- gram has resulted in strong community support, reduced vehicular speeds, and increased yielding to pedestrians in crosswalks. The city works closely with the fire department to ensure that the traffic-calming elements increase overall safety while maintaining the ability of emergency vehicles to respond in a timely manner. To that end, the city reviews each design with the fire department, closely evaluates the locations of the raised devices to ensure a minimal number are located on primary response routes, locates the devices so as to maximize their benefits to pedestrians, and design the slopes to be accessible for the vehicles anticipated on the given street. The slopes of the devices generally range between 5% and 7%. The steeper slopes are used on lower- volume residential streets, and the more gradual slopes are used on busier response routes. Before–after speed studies and resident surveys are the primary methods used to evaluate the effectiveness of traf- fic calming, and are generally conducted within a year of construction. Brattle Street is a two-lane road in a residen- tial area with a posted speed limit of 30 mph. Before–after speed studies involved observing the percentage of vehicles exceeding the speed limit and comparison of 85th percentile speeds (“Brattle Street Traffic Calming Project Evaluation” 2012). The 2010 Brattle Street project used a series of curb extensions, chicanes, and pedestrian crossing islands (Fig- ure 37). As a result of the traffic-calming treatments applied, there was a reduction of the number of vehicles traveling from 25 mph to 30 mph and an increase in the number of vehicles traveling from 21 mph to 25 mph. Observed 85th percentile speeds decreased from 31 mph to 30 mph. Postcards with a link to an online resident survey were mailed to 417 residents postproject, and received at a 13% response rate (“Brattle Street Traffic Calming Project Eval- uation” 2012). A postproject resident survey was also sent to 140 resi- dents after the Windsor Street traffic-calming project, along with a stamped self-addressed return envelope, and received an 11% response rate. The survey asked about resident per- ceptions of improved safety for pedestrians. It also asked for

75 opinions on the overall look of the street, if enough commu- nity outreach was performed, and for opinions of the raised intersections. Overall, 69% of those surveyed believed the overall look of the street improved, traffic speeds were reduced, and pedestrian safety was improved. Only 13% had a negative view of the project. Of those surveyed, 56% did not know if the city did a good job reaching out to resi- dents, 38% believe it did, and only 6% believed it did not. Two-thirds of respondents reported they would like to see more similar projects in the future (“Windsor Street Traffic Calming Project Evaluation” n.d.). Resident survey results provide an excellent tool for learn- ing of community concerns about project implementation and the community’s perceptions of safety treatments installed. Because speed reductions are a primary goal of the city’s traffic-calming projects, and have been shown to significantly reduce crash rates and severity, the before-and-after speed studies provide an excellent metric to evaluate how effective the traffic-calming efforts have been. The city has completed numerous successful traffic-calming projects throughout its street network over the past 20 years and has analyzed overall crash data during those decades (Figure 38). It has found that there has been about a 33% reduction in crashes, and that the severity of crashes has generally been reduced. In short, the speed study and resident survey results together allow city officials to gauge effectiveness in treat- ment selection and project execution, and if improvements can be made for future projects. In addition, the program takes advantage of planned programs to implement improve- ments in a cost-effective way. San Francisco, California—Data-Driven Investment Approach for Achieving Vision Zero Adapted from Achieving Vision Zero: A Data-Driven Investment Strategy for Eliminating Pedestrian Fatalities on a Citywide Level, by Kronenberg et al. (2015) with supplemental information provided by SFMTA In 2014, San Francisco Municipal Transportation Agency (SFMTA) completed a unique data-driven process for identi- fying problematic pedestrian safety locations and matching them with highly effective, yet low-cost countermeasures. The primary goal was to create a systematic, data-driven approach to meet the Vision Zero goal of eliminating fatali- ties in the most cost-effective manner. The process output generated a prioritized pedestrian safety project list of 195 intersections over 5 years that targeted the specific safety needs at each location. This strategy exemplifies a process to help meet the challenge many cities and jurisdictions face when trying to prioritize safety improvement projects with limited funding and resources. City planners, engineers, and analysts created a “new, streamlined approach for capital- constrained project development and prioritization that is applicable to all types of safety improvements across any jurisdiction” (Kronenberg et al. 2015, p.2). Development of the strategy was based on the hypothesis that pedestrian injuries and fatalities are not merely the result of random unpreventable accidents, but are caused instead by specific user behaviors and built environment circum- stances that can be changed. Engineering tools and coun- termeasures proven to elicit safer behaviors and reduce the likelihood and severity of user mistakes were identified for locations in the city where interventions were needed most. First, a thorough analysis of 5 years of network crash data was conducted to identify high-injury and -fatality locations and create pedestrian collision profiles for each location. The project team then examined and assigned costs to potentially effective countermeasures, matched to each collision profile. Finally, a scenario-planning exercise was used to evaluate and select a final pedestrian safety implementation strategy for each location (Kronenberg et al. 2015). Data analysis was entirely dependent on the San Fran- cisco Department of Public Health’s (SFDPH’s) TransBASE data set, which provided an inventory of all pedestrian inju- ries and fatalities caused by motor vehicle collisions between 2007 and 2011. Extensive data on each collision were included in pedestrian injury records along with more than 200 spatially referenced variables collected from multiple other agencies. Each pedestrian injury record and spatially referenced variable was linked to an intersection or street segment to allow spatial analysis. Location-specific network information included the presence of destinations featuring businesses, educational institutions, and community ameni- ties, as well as collision characteristics, infrastructure, traf- fic volumes, zoning, population demographics, and health exposures. Additional land use and transportation data aug- mented the primary data set and were attained from the U.S. Census, San Francisco Planning Department, San Francisco Unified School District, SFDPH, SFMTA, and San Fran- cisco Department of Public Works (Kronenberg et al. 2015). Because the San Francisco jurisdiction is composed of more than 1,100 miles of roadway, a prioritization approach was needed. Therefore, a high-injury network was developed that focused on the most problematic locations. The focus was on identifying high-injury corridors, which were sum- marized according to corridor length, fatalities and severe injuries per mile, total injuries per mile, and total weighted injuries per mile. Eighty-seven corridors were identified that comprised 6% of the city’s street miles, yet 60% of total pedestrian injuries during the period of study. Though intersection-level data were initially aggregated into the high-injury corridors to facilitate effective corridor counter- measures such as road diets and speed control treatments, pedestrian injury accidents that were intersection related were also analyzed to focus attention on particular high- injury intersections (Kronenberg et al. 2015).

76 Each intersection and corridor in the high-injury network needed to be matched with site-specific safety countermea- sure recommendations. To this end, 14 separate collision profiles were created by analyzing common crash-related factors obtained from extensive review of local, national, and international crash studies, and confined to variables avail- able in the combined data set. These factors included person characteristics (i.e., age and gender), collision circumstances (i.e., driver turning left at intersection or pedestrian outside of crosswalk), and environmental conditions (i.e., high traf- fic volumes, time of day, or proximity to land use types). Collision profiles included the related crash factors and a description of the collision type. For example, • The collision profile “Complex Intersection” included the crash factor “five-leg or freeway ramp” and the description, “Collisions that occurred in intersection with complex travel patterns”; and • The collision profile “Mid-Block Collision” included the crash factors “driver failure to yield or pedestrian failure to cross in crosswalk” and “high vehicle vol- ume,” and the description, “Collisions that occurred at midblock locations with or without a crosswalk” (Kronenberg et al. 2015). Of all injuries and fatalities, 90% were captured in the assembled list of 14 crash profiles, which was used to inform the subsequent investment cost analysis and project priori- tization (Kronenberg et al. 2015). Possible safety counter- measures were identified for each collision profile based on known quantifiable measures of effectiveness (MOEs). MOEs included collision reduction, crash modification factors, driver yielding, speed reductions, injury-severity reductions, and increase in compliance. Site-specific needs could then be addressed according to level of expected effec- tiveness of targeted pedestrian safety countermeasures. The list of possible countermeasures included four categories: 1. Signalization: LPIs, PCSs, flashing beacons, RRFBs, PHBs, improved signal timing, scramble phasing, protected left-turn phasing, APS, and adding new traffic signals 2. Geometric: roundabouts, raised crosswalks/speed tables, refuge islands, painted medians, corner bulb- outs, temporary measures, lane narrowing, road diets, new crosswalks, crosswalk removal, chicanes, chokers, and curb ramps 3. Signs, Markings, and Operational: lighting, auto- mated speed enforcement, no turn on red, no left turn, advanced stop lines, daylighting, high-visibility crosswalks, overhead crossing signs, and pedestrian warning signs 4. Speed Control Measures and Miscellaneous: speed humps, portable speed trailer or speed display signs, shared space, channelization, landscaping, lower speed limits, vehicle restriction, and closures (Kro- nenberg et al. 2015). Countermeasure costs were developed to “facilitate cost–benefit analysis of different investment scenarios and to estimate a total cost for the citywide program.” Possible treatments were categorized according to the relative cost of implementing the measure or measures at any given location, according to the following: low = less than $10,000; medium = between $10,000 and $100,000; and high = greater than $100,000. Each countermeasure listed earlier was assigned to a cost category (Kronenberg et al. 2015). The project team then could compare the relative effec- tiveness of each countermeasure against its relative cost, and categorize each into high, medium, and low cost-effective- ness. An example of some pairings is shown in Figure 39. FIGURE 39 San Francisco collision profiles and countermeasure pairings. Source: Kronenburg et al. (2015). The process of creating collision profiles based off the high-injury network crash data and assigning cost- effective countermeasures for each profile generated a large list of potential pedestrian safety improvement proj- ects. An investment scenario and evaluation criteria were developed that allowed the city to evaluate these projects within long-term capital constraints and to understand the tradeoffs involved with each possible approach. Three investment strategy scenarios were developed that estab- lished criteria to evaluate and compare projects and iden- tify the preferred investment scenario. They helped guide the city in project selection and prioritization as part of its pedestrian safety Capital Improvement Program. They are as follows: 1. Location Based: Focuses investments on high injury corridors with the highest concentration of collisions per mile.

77 2. Collision Profile Based: Focuses investments to the locations across the city with the most severe injury collision profiles. 3. Countermeasure Based: Favors countermeasures with the shortest implementation time frame and low- est capital cost (Kronenberg et al. 2015). Crash modification factors (CMFs) were applied to project the reduction of fatalities and severe injuries under each sce- nario. CMFs at intersections with multiple countermeasures were multiplied. Projected reductions in fatalities and injuries under each scenario were also evaluated according to capital cost, time of implementation, and impacts to other modes of travel. Combinations with other nonengineering treatments such as education and enforcement were also considered. Decision makers evaluated all three scenarios and arrived at a hybrid, two-phased selection approach. The first phase prioritized projects with quick and inexpen- sive treatments with large short-term benefits at locations with five collision profiles. Phase 1 treatments included LPIs, temporary medians, parking and turn prohibitions, traffic circles, and reduced lane width. The second phase addressed medium- to long-term projects with more expensive needs and the remaining collision profiles. It was decided that the highest-priority locations were along the high-injury network with the highest weighted count of injuries. This created a list of 195 intersections that com- prised 71% of all fatal and severe injuries along the high- injury network (Kronenberg et al. 2015). SFMTA’s data-driven investment approach for achieving Vision Zero is an innovative process for producing and pri- oritizing a comprehensive list of projects and countermea- sures for crash hot spot areas that address multiple collision types and locations while focusing on fatal and severe injury pedestrian crash reduction. However, it does not take into account current site conditions, or final diagnosis, and there- fore once project selections have taken place, they must be validated using photographs, drawings, site visits, and GIS information in order to create a list of countermeasures that can be realistically implemented (Kronenberg et al. 2015). New York, New York—Vision Zero Pedestrian Safety Plan Information compiled from the survey responses with supplemental information provided by NYC DOT Pedestrian-related traffic fatalities have fallen significantly in New York City (NYC) since 1990 (NYC Vision Zero Over- view: http:/www.nyc.gov/html/visionzero/pages/home/ home.shtml). The city has become recognized both interna- tionally and nationally for its safe street designs and policies. The city attributes engineering solutions and changes at dan- gerous locations for having contributed to a 34% decrease in fatalities since 2005. However, despite these improving statistics, since 2011 an average of 250 fatalities and 4,000 injuries related to pedestrian crashes have occurred per year. Improving pedestrian safety in all five boroughs continues to be a significant challenge. To aggressively address ongo- ing pedestrian safety issues, in January 2014, under Mayor Bill de Blasio, the NYC Department of Transportation (NYC DOT) adopted the Vision Zero framework. The Vision Zero strategy was created and first adopted in Sweden in 1997. As suggested by the name, Vision Zero targets reducing fatalities and serious injuries to as close to zero as possible within a certain time frame. Such a vision alters the thinking toward deciding on road safety projects and locations from a traditional cost–benefit structure that views crashes as unavoidable accidents, to one based on eth- ics, responsibility, safety, and mechanisms for change. The responsibility for road safety is no longer solely that of the user, but shared by the system designers, the overall safety culture, and the built environment, which includes the roadway, planners, engineers, vehicles, road users, emer- gency response, enforcement, multiple city agencies, leg- islation, and policy. NYC’s Vision Zero (and others) views fatalities and serious injuries as system failures that not only can be minimized but can be eliminated through proper system design and shared user responsibility. Where users fail to follow the rules and safely utilize the system, it is the designer’s responsibility to take steps to reduce the probabil- ity of people being killed or seriously injured, by installing the proper safety countermeasures or improving the design of the transport systems. The NYC DOT Vision Zero Action Plan (City of New York 2014) states, “The status quo is unacceptable. The City of New York must no longer regard traffic crashes as mere ‘accidents,’ but rather as preventable incidents that can be systematically addressed. No level of fatality on city streets is inevitable or acceptable.” The plan focuses on three key tenants that guide all decisions: 1. There is no acceptable level of death and injury on streets. 2. Traffic deaths and injuries are not accidents but crashes that can be prevented. 3. The public expects safe behavior on city streets and participates in culture change. As part of the plan, NYC DOT officials have engaged in wide-ranging studies of the problems and comprehensive com- munity outreach in all five boroughs. During this extensive outreach, further insight was attained into problem locations, causes, and the types of safety interventions needed. Through- out the five boroughs, conflicts involving senior pedestrians are disproportionately high. Pedestrian fatalities caused by

78 trucks and large commercial vehicles are also a major problem in Manhattan’s densely packed streets. Throughout NYC, dan- gerous driver choices such as failure to yield and speeding are primary and contributing causes to pedestrian fatalities and injuries. Pedestrian compliance issues and visibility problems also increase safety risks to pedestrians. Feedback from numerous community outreach efforts combined with New York City Police Department (NYPD), New York State Department of Motor Vehicles, and New York State DOT borough-specific crash and safety data has gone toward creating Borough Pedestrian Safety Action Plans. Community outreach efforts included nine public workshops and public comments attained through a Vision Zero interactive input map. This map allows residents to post comments on trouble areas in the street network. More than 11,000 pedestrian safety issues have been identified from public comments on the interactive map. Of the complaints, 21% involved speeding and 21% involved drivers failing to yield to pedestrians. Also, 69% of workshop participants listed wide arterial streets as the first priority for pedestrian safety improvements. Pedestrians killed or severely injured (KSI) density heat maps for each borough were then created based on boroughwide rather than citywide crash data (Fig- ure 40). Results allowed agencies to identify problem loca- tions on a corridor, intersection, and area-wide basis. The Borough Pedestrian Safety Action Plans “pinpoint the conditions and characteristics of pedestrian fatalities and severe injuries; they also identify corridors, intersec- tions, and areas that disproportionately account for pedes- trian fatalities and severe injuries, prioritizing them for safety interventions” (NYC Vision Zero Borough Pedes- trian Safety Action Plans: http:/www.nyc.gov/html/dot/ html/pedestrians/ped-safety-action-plan.shtml). Plan pri- orities include efforts to 1. Significantly expand exclusive pedestrian crossing time on all priority corridors by the end of 2017, 2. Modify signal timing to reduce off-peak speeding on all priority corridors by the end of 2017, 3. Install expanded speed limit signage on all priority corridors by the end of 2015, 4. Coordinate with Metropolitan Transportation Author- ity to ensure bus operations contribute to awareness and safety of pedestrians, and 5. Proactively design for pedestrian safety in high- growth areas including locations in the Housing New York Plan. FIGURE 40 Manhattan Borough Pedestrian Safety Action Plan map. Source: Pedestrian Safety Action Plan: Vision Zero Manhattan (http://www.nyc.gov/html/dot/downloads/pdf/ped-safety-action-plan-manhattan.pdf).

79 In addition to these actions, NYC DOT will to commit to install 50 new safety projects per year and create other new safety initiatives. Because of the specific nature of problems within each borough, community outreach and borough- specific crash data are vital in determining and implement- ing new projects and deciding on treatments. After determining project locations through a combina- tion of crash data analysis and community outreach, NYC DOT selects which safety countermeasures to use at which pedestrian crossings based on pedestrian and vehicular vol- umes, level of service analysis, street width, signal phasing, constructability (drainage and utilities), street network anal- ysis, and community review. Depending on its scope, each project design is reviewed by multiple units within NYC DOT, including signal operations, highway design, sidewalk inspection and maintenance, traffic planning, and borough engineering. Major projects are then sent to multiple outside agencies for review, including the fire department, depart- ment of sanitation, department of parks and recreation, and police department. Common engineering treatments include raised medians, pedestrian refuge islands, curb extensions and bulb-outs, reduced corner radii, road diets, and narrow lane widths (standard is 10 ft; 11 ft is provided for bus and truck routes). Raised medians are used on short block segments where two refuge islands merge to create a continuous median. Pedes- trian refuge islands are typically installed when four- to three-lane road diets are implemented (Figure 41). FIGURE 41 Median islands on Crotona Avenue, Bronx, New York City. Source: NYC DOT Pedestrian Projects (http://www. nyc.gov/html/dot/html/pedestrians/pedestrian-projects.shtml). Curb extensions are, however, used only where they do not interrupt storm drainage or storm water management pro- cesses. Road diets are generally from four to three lanes, and implemented where vehicular volumes are under 700 cars at peak hour per direction. Corridorwide speed calming, neigh- borhood slow zones, and enhanced illumination at crossings are also utilized in all five boroughs. Traffic-calming schemes are generally based around narrowing lanes and road diets and creating wide parking lanes (Figure 42). Leading pedes- trian intervals are installed at high-crash intersections and intersections with high amounts of turning conflicts. FIGURE 42 Traffic calming at NY-62 with road diet, bicycle lanes, and parking lanes with curb extensions and traffic circle. Source: http://www.dot.ny.gov. NYC DOT Vision Zero also seeks to make streets in all five boroughs safer through public dialogue and educa- tion, law enforcement, innovative street design treatments, and legislation. Thus, newer treatments, including innova- tive traffic control devices such as raised crosswalks and intersections, parking restriction through daylighting (i.e., opening up the sight distance at intersection corners by elim- inating parking near the intersection or adding curb exten- sions or both), and turning treatments at intersections meant to reduce truck-to-pedestrian crashes, are being considered. Since before 2014, NYC DOT street design aspects have included more accessible pedestrian signals at high-priority locations; shortened crossing distances (through road nar- rowing, curb extensions, and signal timing) in identified zones with higher concentrations of senior residents; install- ing tactile warning strips for visually impaired pedestrians; providing new crosswalks or crosswalk enhancements at places that have been identified as desired or high-pedestrian crossing locations; clear lane markings; adding additional signalization for turns at intersections with high amounts of turning conflicts; designing spaces for buses, cyclists, driv- ers, and pedestrians; significant lane reductions; and provid- ing less-complex intersection configurations.

80 In 2014 (Vision Zero Year One), NYC DOT implemented numerous engineering projects, which included traffic-calm- ing strategies such as arterial slow zones, neighborhood slow zones, and speed humps. A 25 mph citywide speed limit was passed after initial efforts seeking 20 mph limits (Figure 43), public outreach efforts created a new dialogue around street safety, and NYPD increased its crack-down on dangerous driving with a 42% increase in speeding tickets and 125% increase in failure-to-yield summons. Education efforts addressed city-employed drivers and public campaigns emphasized safer street use (NYC Vision Zero One Year Report April 2015: http:/www.nyc.gov/html/visionzero/ assets/downloads/pdf/vision-zero-1-year-report.pdf). FIGURE 43 Efforts to establish 20 mph speed limit. Source: http://www.nysenate.gov. Arterials comprised only 15% of vehicle mileage, yet accounted for 60% of pedestrian fatalities. This was primar- ily as a result of high vehicle speeds. In response, NYC DOT initiated the Arterial Slow Zone Program, which used a com- bination of treatments and tools (Figure 44). These included lower speed limits, adjusting signal timing to discourage high speeds by coordinating consecutive intersection sig- nals to be consistent with the slower speeds, and increased NYPD traffic enforcement. Signage generally indicates the slow zone, and NYPD regularly employs temporary speed boards and feedback signs throughout the zone. Street design is understood to be most effective when both the public is educated and street safety rules are strictly enforced. To help enforce reduced vehicle speed limits, the city attained authorization from the state to con- tinue its red light camera program and expand the speed camera program. To help educate the public and direct the street design process, the Vision Zero Action Plan also outlined 103 initiatives to be undertaken by numerous agencies, but spearheaded by NYC DOT, NYPD, the Taxi and Limousine Commission, the Department of Citywide Administration Services, and the Department of Health and Mental Hygiene. All initiatives have already begun and many are completed. Under NYC Vision Zero, once a project plan has been proposed, community feedback is a vital part of how NYC DOT decides what safety features to include or to modify from the plan. For example, NYC DOT has recently untaken numerous projects to revamp Queens Boulevard and Bruckner Boulevard, including protected bike lanes in the overall designs. However, in July 2015 public bike advocacy groups raised their concerns about the lack of bike lanes in other projects, such as Riverside Drive, Eighth, Street and Atlantic Avenue. Such concerns from the community are taken into serious consideration as part of Vision Zero (StreetsBlog NYC. July 28, 2015. Accessed at: http:/www.streetsblog. org/2015/07/28/tish-james-calls-on-dot-to-make-bike-lanes- standard-on-vision-zero-projects/#more-348505). FIGURE 44 Arterial slow zone on Northern Boulevard, New York City, New York. Source: http://www.nyc.gov. In summary, project selection through the Vision Zero philosophy starts with the ethic that nearly all serious injury accidents can be prevented through the combination of street design, education, and enforcement efforts. Community outreach and robust data analysis based on borough-specific conditions further helps refine project selection. Redesign- ing high-crash problem locations is more efficient when initiatives for all involved agencies are clearly outlined, allowing greater interagency coordination. Finally, commu- nity feedback on project proposals helps refine treatments selections to best serve all road users, thus facilitating the public cultural change that is part of the Vision Zero method. CASE EXAMPLES—INNOVATIVE TREATMENTS AND APPROACHES FROM OTHER SOURCES Student Waiting Areas and Stand-Back Lines Phoenix, Arizona From Safe Routes to School Guide—Case Examples, Engineering (National Center for Safe Routes to School, retrieved from http:/guide.saferoutesinfo.org/)

81 Unfortunately, many school crossings are at busy streets, and many of the sidewalks in Phoenix were built before the time when sidewalk buffer areas were required as a part of the design to separate pedestrians from motor vehicle traf- fic. It is important to provide a separation between moving vehicles and young children waiting to cross a busy street. This is not possible with a 5-ft-wide sidewalk. One such school crossing was identified by the Washing- ton Elementary School District in northwest Phoenix. This is a crossing for R.E. Miller Elementary School for nearly 100 children over a busy five-lane street with nearly 40,000 motor vehicles per day. Despite the presence of two cross- ing guards and a 15 mph school zone, the school district expressed a concern about the large groups of children wait- ing on a 5-ft-wide sidewalk before crossing. The school district, city, and property owners worked together on a solution to provide a safe area for students to wait. The property owner (church) provided an easement to build a 10 ft by 20 ft waiting area behind the sidewalk (Fig- ure 45). The school district moved the existing wood fence behind the new student waiting pad, and the city modified the landscaping behind the sidewalk, poured a concrete pad for students, and placed a “stand-back” line between the sidewalk and student waiting area. These low-cost and low-tech measures provided a considerable safety benefit at the crosswalk. Since then, Phoenix has built nearly 80 student waiting areas at major crossings where large num- bers of students congregate before crossing. More of the painted stand-back lines have been installed at numerous school crossings. This example illustrates that a jurisdiction does not have to spend a lot of money to obtain a big safety divi- dend. Some of the least-expensive measures can have a big impact on safety. Traffic Control Plan Sabin Elementary, Portland, Oregon From Safe Routes to School Guide—Case Examples, Engineering (National Center for Safe Routes to School, retrieved from http:/guide.saferoutesinfo.org/) As a result of a technical analysis of pedestrian safety around Portland Schools, the city traffic-calming program identified Sabin Elementary as a high priority for traffic- calming measures. The school, with 500 children enrolled, is served by a tra- ditional grid street pattern with north–south and east–west arterials. Some of the problems affecting the area included traffic congestion during pick-up and drop-off times and unsafe pedestrian crossings. FIGURE 45 Images of student waiting area in Phoenix, Arizona. These images highlight the differences before and after a waiting area and stand-back line were installed at Richard E. Miller Elementary School in Phoenix. Source: Mike Cynecki and Safe Routes to School Guide case study. In order to start the planning process, the city staff, along with key stakeholders, created the Sabin School Safety Com- mittee. Among people involved in the committee were the school principal, community members, the school PTA, and Portland police. The committee identified major problems to work on, developed goals, and adopted the specific objectives of decreasing speeds on the north–south arterials, increasing visibility at key intersections, and improving crossing safety at two east–west arterial streets. Community feedback on the pro- gram was gauged through an open house, as well as a survey distributed to properties near the proposed improvement sites. Once community concerns were addressed, the city con- ducted a pilot program of temporary improvements to see how they would affect driving conditions in the area. The pilot pro- gram demonstrated positive traffic-calming and pedestrian safety effects, and the city went forward with the final perma- nent program. The total cost for the project was $54,000. After the traffic-calming measures were made perma- nent, two-way car conflicts have been reduced and pedes- trian safety has increased. Discussions with nearby residents indicate that traffic congestion has decreased, as have dan- gerous conditions during the opening and closing of school (Zegeer et al. 2013). 15 mph School Zones Arizona From Safe Routes to School Guide—Case Examples, Engineering (National Center for Safe Routes to School, retrieved from http:/guide.saferoutesinfo.org/) Arizona has been using 15 mph school zone crossings for elementary and middle schools since 1950 when the legisla- ture passed State Law ARS 28-797 allowing for the uniform application of these special reduced-speed zones. School zones are operated by the schools and are established with roll-out 15 mph speed limit signs and portable “Stop When Children in Crosswalk” signs during student arrival and dis- missal times (Figure 46). Their use is limited to crossings where there is not the benefit of a traffic signal or stop sign at the crossing. The law requires an engineering study before

82 establishing the 15 mph school zone and the school must sign a written agreement with the local traffic authority to operate the zone. These special crosswalks are marked with yellow lines and the portable signs are typically allowed in the street 45 minutes before the start of school and 30 minutes after school dismissal. The 15 mph school zone portable signs are typically removed from the street during noncrossing times, but are allowed in the street during the entire school day if the crossing is abutting the school grounds. The law also strictly prohibits passing on the approach to the crossing, and the 15 mph speed limit extends from the advance “15 mph School in Session” portable sign through the crosswalk. The 15 mph school zones cannot be used within 600 ft of a traffic signal, stop sign, or another 15 mph school zone crossing. FIGURE 46 15 mph slow school zone in Phoenix, Arizona. Source: Mike Cynecki and Safe Routes to School Guide case study. Also see “Elementary School Crosswalk Enhancement Pro- gram in Bellevue, WA,” a case study prepared for PEDSAFE, available at http:/pedbikesafe.org/PEDSAFE/casestudies_ detail.cfm?CS_NUM=30&op =L&subop =I&state_ name=Washington. Speed-Sensitive Traffic Signals Boulder, Colorado; Arlington, Virginia; and Washington, D.C. From National Safe Routes to School Clearinghouse website http:/guide.saferoutesinfo.org/engineering/speed_ sensitive_signals.cfm High-speed motor vehicles pose a serious threat to the safety of children who are crossing arterial streets near schools and are one of the largest challenges in providing safe routes to school. Innovative measures have been used to reduce motor vehicle speeds such as the speed-sensitive signals used in Boulder, Colorado; Arlington, Virginia; and Washington, D.C. The signals use pavement loops to detect the speed of a motor vehicle. If the motor vehicle exceeds the speed limit, the traffic signal ahead displays a red light. Drivers learn that speeding on such streets will require them to stop at the light and be further delayed. The sign “speed-sensitive signal” conveys that message to drivers. Speed-sensitive signals were added as part of a traffic- calming project at Sunnyslope High School in Phoenix, Ari- zona. The signals monitored motorists’ speed and flashed their driving speed and a bright LED strobe light if a motor- ist exceeded the 35 mph speed limit by 5 mph or more. Along with other traffic improvements, including a stag- gered crosswalk, the speed-sensitive signals resulted in slower motorist speeds (on average cars went 6 mph under the speed limit) and a reduction in pedestrian crashes. The avenue averaged 32 pedestrian crashes per year during the previous 3 years, but only one crash in the 6 months follow- ing the project (Zegeer et al. 2013).