Enhancing Pedestrian Volume Estimation and Developing HCM Pedestrian Methodologies for Safe and Sustainable Communities (2022)

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.

230 Appendix B: Stakeholder Interviews Introduction This appendix presents the feedback received from stakeholder interviews conducted by the research team. Organizations to be interviewed were proposed by the research team in the Amplified Work Plan and were selected to provide a mix of agency types: state DOTs, MPOs, and cities and counties. The selected organizations represent a range of geographic locations and population sizes, and are known as leaders in improving pedestrian transportation. In addition to the public agency interviews, representatives from several TRB committees were interviewed, including the committees on HCQS, Pedestrians, and Accessible Transportation and Mobility. These interviews provided additional points-of-view, including consultants and academic practitioners, and also set the stage for continuing dialog between the project and the committees. In particular, the HCQS Committee will be involved in reviewing and approving any new material for the HCM once the project is completed. Draft interview questions were provided for panel comment in the April–June 2018 Quarterly Progress Report. Panel comments on the interview questions were received on August 9 and the project team revised the questions in response to the panel comments. The revised questions, the interview plan, and documentation that the research team’s interviewers had gone through the Stage 1 IRB Social Basic training course were submitted to Penn State University’s IRB. An Exemption Determination was received on August 30, 2018 from Penn State’s Office of Research Protection, finding that the proposed interviews did not require formal IRB review. While waiting for panel comments and IRB approval, the project team identified suitable contacts at each of the identified transportation agencies or TRB committees that had been proposed to be contacted in the project’s Amplified Work Plan. These contacts were identified based on personal knowledge of research team members, use of the Association of Pedestrian and Bicycle Professionals directory, and agency websites. Once the IRB exemption was obtained, the project team proceeded to make contact with the 63 organizations and individuals identified to be interviewed. Information from 40 interviews are summarized in this appendix. The remaining potential interviewees declined to be interviewed or failed to respond to multiple contact attempts. Table B1 lists the organizations represented in the summary results presented in this appendix. Because the interviewee selection process was not intended to produce a set of interviewees representative of a group (e.g., state DOTs) as a whole, results are generally provided in aggregate form. However, results intended to help inform the decisions about future project activities (in particular, the selection of Task 6 research activities) have been broken down by organization type.

231 Table B1. Organizations Interviewed by Type. State DOTs Caltrans Colorado DOT Florida DOT Louisiana DOTD Maryland DOT Massachusetts DOT Michigan DOT Minnesota DOT North Carolina DOT Oregon DOT South Dakota DOT Metropolitan Planning Organizations Compass (Boise, ID) Boston (MA) Region MPO Memphis (TN) Urban Area MPO Palm Beach (FL) Transportation Planning Agency Delaware Valley RPC (Philadelphia, PA) Metro (Portland, OR) Richmond (VA) Area MPO Southern California Ass’n. of Governments (Los Angeles, CA) Metropolitan Transportation Commission (San Francisco, CA) Cities and counties Boulder, CO Charlotte, NC Clackamas County, OR Denver, CO New York, NY Orlando, FL Portland, OR Washington, D.C. TRB Highway Capacity Committee 6 members of the committee leadership (5 consultants, 1 state DOT) TRB Pedestrians Committee 4 members of the committee and section leadership (3 academic practitioners, 1 consultant) TRB Accessible Transportation & Mobility Committee 1 member of the committee leadership (included with Pedestrian Committee summary statistics) Panel members not included above 1 member (included with Ped. Committee summary statistics) Interview Results This section provides a summary of the interview results, organized in order of the questions asked during the interviews. In many cases, interviewees provided multiple responses to a single question (e.g., uses for count data); therefore, the number of individual responses often exceeds the number of interviewees.

232 Pedestrian Volume Counting Pedestrian Volume Data Collection A total of 29 interviewees responded that they collect pedestrian volume data:  MPOs (7)  DOTs (8)  Consultants/academics (7)  Cities and counties (7) Of these, ten responded that have developed standard methods for counting pedestrians, while four mentioned that they are in the process of doing so. Researchers obtained and reviewed standard methods from those agencies that had developed one. The National Bicycle and Pedestrian Documentation Project guidelines were specifically mentioned by one agency. Another agency mentioned Chapter 4 of the 2016 Traffic Monitoring Guide and information related to nonmotorized counting. The interviewees who did not directly collect volume data worked with data collected by others. Counting Methods and Technologies The following methods and technologies are used by those interviewees who collect pedestrian volume data:  Manual. Fifteen interviewees use this method and expressed mixed reviews: some are satisfied, while others are not satisfied due to reduced accuracy with higher pedestrian volumes and cost. One stated advantage of this method was the ability to count crossings at different midblock locations, as individuals’ fields of view were greater than what a single camera could achieve.  Automatic from video. Twelve interviewees use this method, with nine using one vendor, one using another vendor, and two did not state. Six interviewees are mostly satisfied with this method and believe the count accuracy is very good; however, two interviewees are dissatisfied due to concerns about accuracy. Sated drawbacks of this method were not being able to conduct directional counts and that it was not feasible to store all the raw video data.  Manual from video. Nine interviewees use this method and are satisfied with it. Stated advantages were its ability to provide directional counts as well as more-detailed information such as counting groups of pedestrians or the presence of children. The time required to develop counts from video, cost, and the ability to count only what was visible within the camera’s field-of-view were stated as drawbacks.  Infrared/pyroelectric. Seven interviewees use this method and are satisfied to somewhat satisfied with it. Advantages that were mentioned included portability and some devices’ ability to provide directional counts. One interviewee manages a program to train jurisdictions on performing pedestrian counts and loans pyroelectric devices to them to perform the counts. Drawbacks that were mentioned included finding appropriate sites to install counters, issues with counting bicyclists along with pedestrians, potential vandalism, and expense.  Piezoelectric: Four interviewees use this method and are satisfied with it. The device’s portability was mentioned as an advantage.  Radio beam: One interviewee used this method and was not satisfied with it. It was expensive, hard to install, required units on both sides of the facility, and was difficult to upload data. On the other hand, it was able to differentiate between pedestrians and bicycles.  Combination: Three interviewees used combination sensors that used passive infrared to count all facility users and inductive loops to count only bicycles; of these, one had just installed their first

233 device and were just gaining experience with it. One interviewee uses a combination of an infrared detector and piezoelectric loop, but has found undercounting with bicyclists. Types of Facilities Counted The following facility types were counted by interviewees:  Intersections and crosswalks (22)  Shared-use paths (19)  Sidewalks (19)  Bridges and tunnels (3)  Stairways (often leading to bridges) (3)  Locations without sidewalks (1) Count Directionality Twelve interviewees typically conducted directional counts, while 19 interviewees typically counted total volumes or did not specify that they conducted directional counts. Count Site Selection The following methods were mentioned as being used to select count sites:  Project-based (sometimes as part of a vehicle-focused intersection count) (16)  Established counting program, with sites selected to obtain a good distribution of locations (6)  Locations with high pedestrian volumes (5)  External input (e.g., citizens, council members) (4)  Before/after studies of safety treatments (3)  Traffic signals (i.e., push-button activations) (2)  Crash history (2)  School sites (1)  Statewide survey targeted at planning directors, engineers, planners, designers, safety specialists, and advocacy groups (1) Count Duration and Frequency The following short-term count durations were used at a typical location:  Peak 1–3 hours (a.m., p.m., sometimes midday) (18)  Twelve to thirteen hours (3)  One week (8)  Two weeks (1) Count frequencies mentioned by interviewees were:  Seven 2-hour periods (a.m., p.m., and weekend on different days), every 2 years (manual)  Three 1-hour periods (a.m., midday, p.m.), every 3 years (automated from video)  Congested intersections: 13-hour counts Thursday–Sunday, every 4 years (automated from video)  One 24-hour period, annually (manual from video)  One week, every 1–2 years (passive infrared)

234  Fifteen to 30 days, annually (passive infrared)  One- to three-month intervals (passive infrared)  Two or three 24-hour periods (Tuesday, Thursday, possibly weekends)  Two 2-hour periods (a.m., p.m.), annually (manual)  Three 2-hour periods (a.m., midday, p.m.), annually (manual)  2 hours a.m., 3 hours p.m., and 2 hours Saturday, every 2 years (manual)  One or two 12-hour counts (passive infrared)  May collect multiple peak-period counts for vehicles and pedestrians on different days, when the counts are used for signal timing projects (manual from video) A few interviewees indicated that they favored counting more sites over counting the same site more than once in the same year (for short-term counts). One agency has been counting for 6 years and has only duplicated sites a few times. Another agency is considering rotating short-term count sites around so that each site is counted once every 3 years. A third has counted every signalized intersection (roughly 1,600) in its jurisdiction in the last 5–7 years. Use of Adjustment Factors Most interviewees do not make adjustments to the raw count data, with the following exceptions:  Adjustments for device-counting errors (e.g., occlusion). Seven organizations that perform automated counts make these adjustments. Two make vendor-specific adjustments. Two calibrate their devices yearly or every 2–3 years. One interviewee mentioned that the decision to adjust depends on the time of day, and is performed using manual observations. One mentioned that video camera data are manually extracted to compare to the automated count results to obtain correction factors.  Converting short-term to long-term counts. Two interviewees responded that their organizations do this; one of these was not sure of the procedure, while the other only does it for shared-use paths and mixed-mode counts. One interviewee noted they plan to do this in the future.  Accounting for future growth. Only one organization responded that they do this, based on regional count data projected into long-term growth. One organization mentioned that they are considering this.  Other. One organization uses seasonal adjustment factors for its pedestrian counts. Another adjusts for special events, holidays, and weather. In addition, one organization mentioned that they convert elderly and child pedestrians into equivalent adult pedestrians using a conversion factor of 1.33. This is done when evaluating sites for installing a marked crosswalk. Count Applications The following count data applications were mentioned by interviewees:  Monitoring facility usage (trends), forecasting, justifying infrastructure investments (15)  Safety projects (e.g., exposure, corridor safety) (14)  Prioritizing pedestrian-specific improvements (10)  Motor vehicle-focused studies (e.g., automobile LOS evaluations) (5)  Traffic signal timing (5)  Local demand (4)

235  Before-and-after studies (3)  Planning (e.g., new trail locations, pedestrian plans) (2)  Crosswalk warrants (2)  Modeling (e.g., activity-based modeling) (1)  Policy-level trend documents (1)  Determine maintenance needs (1)  Design crosswalks and intersections (1)  Plan construction detours (1)  Mode split determination for trip generation studies (1) Pedestrian Safety Countermeasures Implemented Countermeasures At intersections, the following countermeasures have been implemented by the interviewees, listed in descending order of use:  Change signal timing (19) – LPI (15) – Exclusive pedestrian phase (7) – Pedestrian push button (3) – Crossing time extender (3) – Automatic pedestrian phase (1)  Curb extensions/bulb-outs (12)  Pavement markings (11) – Reposition stop lines (4) – Yield lines (shark teeth) (1)  Crosswalk treatments (10) – Raised crosswalks (4)  Pedestrian countdown timers (9)  Advanced signage (7)  Pedestrian refuge islands (6)  Left-turn treatments (4) – Flashing yellow arrow (e.g., use pedestrian inhibit feature) (2) – Lag protected left turns so pedestrians are into or out of the conflict area (1) – Prohibit left turn (1) – Left-turn speed calming (vertical delineators) (1)  Pedestrian hybrid beacon/HAWK (4)  RRFB (4)  Prohibit right-turn-on-red (4)  Audible signals (2)  Roundabouts (2)  Curb radius change (2)  Porkchop island (2)  Lighting improvements (2)  Gateway treatments (1)

236  Pedestrian signal (1)  Shorter crossing distance with road diet (1)  Establish target speeds (1) At midblock crossings, the following countermeasures have been implemented by the interviewees, listed in descending order of use:  Beacons (19) – RRFB (12) – Pedestrian hybrid beacon/HAWK (11) – Static beacon/yellow flashing beacon (3)  Signage (in-street crosswalk signs, advance signage, warning signs) (10)  Curb extensions/bulb-outs (8)  Median refuge (8)  Midblock pedestrian signals (5)  Midblock crosswalks (5)  Pedestrian-activated signals (3)  Improved pavement markings, raised pavement markers (3)  Yield lines (shark teeth) (2)  Crossing illumination (2)  Raised crossings (1)  Offset crossing (1)  Overpass/underpass (1)  High emphasis crosswalk (1)  Gateway effect (1) Along street segments, the following countermeasures have been implemented by the interviewees, listed in descending order of use:  Introduce or improve sidewalks or separated pathways (11)  Road diet (e.g., 4-to-3 cross-section reduction) (7)  Landscaping and/or amenities (e.g., trees to increase pedestrian sense of safety and/or provide shade; benches, plazas, and shard spaces) (6)  Buffers (4)  Signage (3)  Increase crossing opportunity frequency (2)  Traffic calming (2)  Separation of bicyclists and pedestrians on multi-use paths (2)  On-street parking (2)  Static beacons (1)  Speed humps (1)  Pedestrian islands on protected bike lanes (1)  Wide shoulders (1)  Posted speed reduction (1)  Narrow lane widths (1)  Speed management (1)

237 Countermeasure Effectiveness Only five organizations had tried to quantify effectiveness of at least one of their safety countermeasures through before-and-after studies using crash data. Therefore, the results in this section are primarily anecdotal. A general rule of thumb provided by one organization was that any countermeasure that increases pedestrian visibility and separation from traffic is generally assumed to be effective. One organization noted that they sometimes base decisions using CMFs. Issues with lack of enforcement (both for motorists and for pedestrians) were mentioned as a major reason for safety countermeasures not being effective. One organization mentioned that countermeasures were generally ineffective along truck routes. Several interviewees did not feel comfortable answering this question due to a lack of quantitative data. Countermeasures rated as somewhat or highly effective (in the right application) were:  Pedestrian hybrid beacon/HAWK (5)  Curb extensions/bulb-outs, but caution has to be used when bike lanes are present (4)  LPIs (4) – one study in a large city showed a 56% reduction in fatal and severe injury crashes after implementation  In-street crosswalk signs (2)  Changing signal timing (2)  Road diets (2)  Exclusive pedestrian phase, in areas with high pedestrian activity (1)  Landscaping (1)  Adding sidewalks and separated paths (1)  Relocating stop bars (1)  Adding crosswalks and ADA ramps, even if pedestrian volumes are low (1)  Raised crosswalk (1)  Static beacon (1)  In-road YIELD TO PEDESTRIANS signs (1)  Lighting (1)  RRFB (1) Countermeasures rated as minimally effective were:  Standard crosswalk markings (2) – particularly across more than two travel lanes  Yield lines (shark teeth) (1) Countermeasures with mixed experience were:  Median refuge islands: three organizations considered these highly effective, but they did not work well on 5-lane streets for one organization because they did not increase pedestrian visibility enough  RRFBs: two agencies considered them highly effective, while three others thought they had varying effectiveness, with high-speed locations and situations where pedestrians were not visible to motorists being situations where they were minimally or not effective  Advanced signage: one agency thought it was highly effective, while another thought it was the least-effective option (although it still helped)

238 Process for Identifying Locations Requiring Countermeasures Organizations used the following to identify locations requiring pedestrian safety countermeasures:  Safety planning and monitoring programs (22) – Crash analysis (10) – Vision Zero (5) – High-injury network analysis (4) – Safe routes to schools (2) – High-crash corridors (1) – Bicycle/pedestrian plans (1) – Systemic safety analysis (1) – Multimodal safety assessments (1) – Sidewalk gap analysis (1)  Public feedback, local community, local agencies (11)  Land use (e.g., new development, high pedestrian areas, near transit (3)  Equity (3)  Requirements (2) – State requirement that all signalized intersections include a pedestrian evaluation (1) – ADA requirements for sidewalks (1)  Add-on to existing project (2)  ActiveTrans tool (1)  Sidewalk and crosswalk gaps (1)  Uncontrolled pedestrian crossing guidelines (1)  Pedestrian demand model for arterials and collectors (1)  Internal cross-disciplinary committee (1)  Staff in the field for other purposes (1)  Traffic signal optimization program (1)  FDOT Traffic Engineering Manual and Design Manual (1) Process for Identifying Suitable Countermeasures The interviewees used the following processes to identify a specific countermeasure once a candidate location had been identified:  Project/site-specific (e.g., review field conditions or crash types and select an appropriate treatment) (9) – Engineering judgement (2)  Guidance documents (e.g., for pedestrian crossings) (4) – Charles Zegeer reports for crosswalks (2) – NCHRP Report 562 worksheet (1) – FDOT Traffic Engineering Manual Section 3.8 (1)  Cost (2)  Built environment factors (2)  CMFs (2)  GIS shapefiles for specific countermeasures (1)

239 Process for Prioritizing Countermeasures The interviewees used the following processes to prioritize the implementation of pedestrian safety countermeasures:  Crash data (13) – Risk model, risk factors (2)  Land use (e.g., transit use, schools, heavy bike volumes) (9)  Pedestrian volume (3)  Community feedback (3)  Based on existing projects (2) – Increase countermeasure scale at a location if needed (1)  Cost (2)  Risk factors (2)  Fill in network gaps (1)  Vehicular volume (1)  Equity (1)  For HAWKs: street class, street volume, distance from nearest signal (1)  Local jurisdiction support (1) One interviewee noted that small projects (less than $5,000) could be handled by the agency’s maintenance staff, while larger projects must go through a more involved funding and design process. Challenges Faced in Implementing Countermeasures Interviewees reported the facing the following challenges when implementing countermeasures:  Cost/budget/insufficient staff (14)  Impact on vehicles (e.g., LPIs increasing driver delays), stakeholder disagreements (9)  Physical constraints (e.g., lack of road space or right-of-way, requirement to maintain 12-foot lanes, existing utilities) (6)  Lack of political will, project- or program-specific controversy (5)  Liability concerns (2)  Difficult to justify in low-demand areas (2)  Lack of engineer awareness/knowledge (2)  Lack of data (2)  Low pedestrian crash numbers (1)  Overuse (1)  Crosswalk volume warrants (1)  Too much reliance on crash data (1)  Availability of research to support the installation of a specific countermeasure (1)  Statewide consistency in applying countermeasures (1)  Developing maintenance plans, maintaining supplies/parts (1)  Identifying where to start along high-crash arterial corridors (1)

240 One organization noted that seemingly simple projects, such as filling a sidewalk gap, can turn out to be very expensive, due to the need for water drainage and detention facilities, and that intersection projects were also expensive. The agency also noted that they had started bringing public health factors into the discussion to help justify installing countermeasures. Analysis Method Improvements to Support Countermeasures The interviewees made the following suggestions for analysis method improvements that would help them identify pedestrian safety needs, select countermeasures, prioritize pedestrian safety projects, and quantify the effects of countermeasures:  More data to quantify the effectiveness of safety countermeasures (14) – More CMFs (4) – Trade-off information (e.g., right-turn channelization could improve vehicular operations but degrade pedestrian safety (1) – More predictive analyses (1)  Pedestrian exposure data (6) – Latent demand models (2) – Standardize exposure methods (1)  Including all factors in risk models, not just crashes (3)  Authoritative guidance (3) – On the level of the Green Book or MUTCD (1) – Clearly listing all impacts of a given countermeasure (1) – Generally (1)  HCM methodologies for all-way stops and roundabouts (2)  Crash data (2) – Integrating EMS/hospital data (1) – Standardized crash data (1) – Better reporting of crash data (1)  Weighting system to improve pedestrian project competitiveness in terms of safety/number of crashes, or consideration of vulnerable users (2)  Integrating safety into LOS methodology (1)  Integrating counts with road network (1)  Estimation model for SRS2 (1)  Research on buffered bike lanes next to pedestrian facilities (1)  Research on impacts of bicycle and scooter use on sidewalks on pedestrian safety (1)  Trade-offs prioritizing deficient sidewalks vs. sidewalk gaps (1)  Trade-offs between pedestrian crossing improvements and traffic delays (1)  Systemic approaches to pedestrian projects (1)  Pedestrian interactions with autonomous vehicles (1) One interviewee commented “With exposure, observed, and predicted severe pedestrian crashes, I think we’ve landed on a good analysis method to identifying pedestrian safety needs and prioritization schemes.”

241 Pedestrian Quality of Service Interviewees were told that pedestrian QOS refers to a pedestrian’s perceived experience using a transportation facility. It can encompass a range of factors, including pedestrian facility infrastructure and amenity quality, level of crowding, and perceived safety and security, among others. Based on this definition, only 12 interviewees indicated that they measure, estimate, or forecast pedestrian QOS (4 MPOs, 3 DOTs and 5 consultants). Reasons given for not measuring QOS were:  Not sure it is useful/necessary (e.g., thought to be most applicable to larger communities, existing street standards already provide the desired comfort level for pedestrians) (6)  Lack of data (3)  Not familiar with it (3)  Not satisfied with HCM methodology (e.g., can lead to counterintuitive results for road diets, focuses too much on crowding) (3)  Difficult to quantify (1)  Don’t measure it, but qualitatively talk about it (1)  It is not required (1)  Low pedestrian volumes (1)  Difficult to use (1)  Staffing limitations (1) One agency indicated that they do not directly estimate pedestrian QOS, but consider crowding and yielding rates in their analysis as a proxy. Additionally, they measure vehicle turning speeds and are considering exposure as a means of measuring pedestrian QOS. Another considers QOS anecdotally. Purposes for Measuring QOS For those who do measure pedestrian QOS, the following purposes were identified:  Alternatives analysis (e.g., shared path analysis, adding pedestrian accommodation to intersections, multimodal arterial study, active transportation plan, one-way to two-way street conversion study) (6)  Understand existing conditions (2)  Prioritize site improvements (2)  Complete Streets program (1)  Implement in software (1)  Teaching HCM methods (1)  Produce map with visual guidance on which roadways may need additional infrastructure (1)  Identify routes for inclusion in the community pedestrian network (1) One interviewee indicated that they are starting a new project to determine the low- and high-stress roads for walking on a network level.

242 Methods Used to Measure QOS Interviewees identified the following methods that they use to measure pedestrian QOS:  HCM (9) – Shared-use path method (2) – Facility method (1)  Florida DOT method (4)  Charlotte’s index-based system (2)  GIS (e.g., auto speeds, auto volumes, number of lanes, pedestrian crashes, sidewalk completion, tree canopy, signalized crossing spacing, connectivity, commercial land use) (2)  CMFs (1)  Oregon DOT PLTS (1)  Pedestrian LOS (Landis et al.) (1)  Software programs (1)  Pedestrian report card (1) Satisfaction with Existing Methods Highway Capacity Manual Most interviewees who had applied the HCM pedestrian methods (a number of them affiliated with the TRB Highway Capacity Committee) were somewhat satisfied with the HCM methods. Two people cited that it was useful to have a procedure that raises awareness of pedestrian needs that could be included in projects or corridor analysis. Challenges identified with the methods consisted of:  Requires too much data, some data not available (3)  Too quantitative and complex (3)  May misrepresent locations with partial sidewalks as results look better than field experience (1)  Missing some aspects of the pedestrian experience (e.g., aesthetic/environmental factors) (2)  Route choice not considered (1)  Can produce counterintuitive results (e.g., for midblock crossings) (1)  Does not measure the impacts of LPIs, medians, sidewalk crowding (1)  Difficult to intuitively understand what the LOS ratings correspond to (1)  Difficult to compare sidewalks and a separated shared-use path (1)  Difficult to apply the method to a sidewalk on the left side of a one-way street (1) One agency was very dissatisfied with the HCM methods, because they felt it does not measure (or accurately measure) the impacts of LPIs, medians, or sidewalk crowding. They felt it does not take into account pedestrian experience, does not accurately reflect what pedestrian delay means, does not take behaviors into account (e.g., how it changes risk-taking or route choice), nor is it dynamic in terms of pedestrian density. That being said, the agency reports vehicular LOS and would like to be able to report pedestrian LOS alongside it if they could be comfortable with what they were reporting. Florida DOT Method Those who had used the FDOT method were somewhat satisfied with it. Some thought it was easy to do whole projects quickly and data input was simple, while others thought that the results appear to be too optimistic and that the method requires a lot of data.

243 Oregon DOT Pedestrian Level of Stress Method ODOT was very satisfied with its pedestrian level of stress method. The variables are easily quantifiable and include ADA factors. It was more accepted by staff planners than the HCM method, and it can be used for any kind of pedestrian facility, unlike the HCM methods. Charlotte DOT Method Two organizations who had used Charlotte’s index-based method were somewhat satisfied with it. Its flexibility was stated as a big advantage: it was thought to work well for multiple intersection types and can accommodate a large variety of possible improvements (e.g., signal timing, delay, pedestrian refuges). However, it was thought to require a lot of data. Interestingly, City of Charlotte staff indicated that the method had been developed in 2007 to support the development of their urban street guidelines related to signalized intersections and that they do not use it much themselves anymore. Portland Metro Method Metro was somewhat satisfied with its GIS-based method. The results were intuitive and matched expectations, it helped identify areas for prioritization, and could be used to develop a helpful summary matrix of the results. However, the analysis was thought to be too high level and broad. Awareness of HCM Methods Among Those Not Using Them Those who did not use HCM methods were asked whether they were aware of the HCM’s methods for sizing crosswalks to accommodate a given pedestrian demand. Responses included:  Familiar with them, but have never used  Density descriptions are from an engineering perspective, and not a planning perspective (i.e., more pedestrians is bad)  Typically work in areas with low pedestrian volumes, so existing sidewalk standards work fine  Not aware of them  Aware of sidewalk methods, but not crosswalk methods. Those who were familiar with the HCM’s QOS methods for streets and intersections were asked why they did not use them. Responses included:  The method shows very little reaction to proposed improvements (e.g., LOS stays at C, or at best goes up one LOS). The thresholds and parameters need to be changed—it doesn’t help make the case to a client to make the improvement  Geared towards high-density urban areas  Pedestrian LOS is not required as a part of ordinance and state regulations  The level of detail is not required by clients

244 Other Thoughts on Pedestrian QOS Some interviewees offered additional thoughts on pedestrian QOS not covered in the interview script:  If facilities are built, we should not just evaluate them from the operational standpoint, but from the viewpoint of how people perceive them for walking. This could help in creating design standards.  It may be a good idea to marry safety performance functions with LOS. LOS should have something to do with safety performance.  Needs not addressed by current methods: – Determining latent demand (2) – Standardized maximum distance between crosswalks (2) – Factors that crate a friendly walking environment (e.g., crossing 5 lanes of travel and physical obstructions is not friendly) – Issues associated with crossing vehicle turning lanes – Impacts of colored pavers on pedestrian QOS  Some indicators for good QOS: – Time to travel – Number of pedestrians per sidewalk (e.g., crowding) – Combination of accessibility and level of use Prioritizing Potential Research Topics The interviewees were told that NCHRP Project 17-87 has budget to research the effects of pedestrian safety countermeasures and pedestrian operations improvements on pedestrian QOS. However, the project will not be able to research every possible topic. The interviewees were given a list of potential research topics and were asked to rate them from 1 (lowest) to 5 (highest) in terms of the priority they would give the topic for being the subject of research. Table B2 presents the results.

245 Table B2. Interviewee Ratings of Potential Research Topics. Topic Overall State DOTs MPOs Local HCQS Cmte. Pedestrian Cmte. n 36 10 8 7 6 5 Effects of signal timing changes on ped QOS 4.1 (13) 3.9 (2) 4.5 (4) 4.1 (3) 4.0 (1) 4.1 (3) Effects of physical safety improvements on ped QOS 4.4 (21) 4.3 (5) 4.3 (5) 4.0 (3) 4.3 (3) 5.0 (5) Evaluating pedestrian QOS crossing a street 3.9 (16) 3.9 (4) 4.1 (4) 3.8 (3) 3.7 (2) 4.4 (3) Evaluating pedestrian QOS walking along a street 3.1 (3) 3.3 (1) 3.3 (1) 3.2 (0) 2.7 (0) 3.2 (1) Determining the pedestrian volume at which pedestrians start walking out of the intended pedestrian path 2.7 (5) 2.4 (2) 3.0 (1) 3.1 (1) 2.5 (1) 2.6 (0) Determining the required usable pedestrian sidewalk or path width for a given pedestrian volume 2.9 (3) 2.3 (1) 3.3 (0) 3.6 (1) 2.7 (1) 2.8 (0) Determining the required crosswalk width for a given pedestrian volume 2.8 (3) 2.6 (1) 3.3 (1) 2.7 (0) 3.2 (1) 2.4 (0) Determining how crosswalk configurations and motorist behaviors affect pedestrian QOS 3.7 (11) 4.3 (4) 3.9 (3) 4.2 (2) 2.7 (0) 3.5 (2) Extending current HCM pedestrian LOS methods to cover missing intersection types 3.6 (9) 3.8 (3) 3.6 (1) 3.0 (0) 4.8 (5) 2.4 (0) Systemwide pedestrian connectivity and its relationship to pedestrian QOS 4.0 (18) 4.1 (5) 4.6 (7) 3.0 (1) 4.2 (3) 3.8 (2) X.X (X) = average rating (number of “5” ratings) Green shading: Group rating was ≥0.5 points higher than the overall rating Red shading: Group rating was ≥0.5 points lower than the overall rating HCQS = Highway Capacity and Quality of Service, QOS = quality of service, DOT = department of transportation, MPO = metropolitan planning organization, Cmte. = Committee Other potential research topics mentioned by interviewees were:  Pavers and other materials that improve the facility aesthetics  Developing simple methods  Developing methods to be adapted into the HCM planning guide  Safety effectiveness in terms of actual safety rather than perceived safety  Relationship between intersection treatments and pedestrian volumes

246  Methods to determine if sidewalks are needed on both sides of the roadway in low-density areas  Quantifying vehicle–pedestrian interactions at signalized intersections  Transit access quality  How to collect pedestrian data in rural settings  Need more data on curb extensions  E-scooter effects on pedestrians  Impacts of speed and speed limit on pedestrians  How to educate public on pedestrian safety  Determining latent demand  How to make an argument to build pedestrian facilities in suburban areas when pedestrian volumes are low, construction costs are relatively high, and many areas have limited facilities  How to use existing data for estimating volumes  Climate impact on pedestrian safety (e.g., need for shade)  Understanding benefits of countermeasures sooner (e.g., not necessarily having to wait for multiple years of crash data) Reasons given by interviewees for rating certain topics as high priorities were:  Safety is the most important indicator for pedestrian QOS (6)  Need methods to justify the importance of pedestrian projects (2)  Interested in signal timing – lack of knowledge about impacts of short cycle lengths and exclusive crossing phases (2)  Completeness of evaluation, addressing missing intersection types in HCM methods (2)  These are things that we deal with most often in our work (2)  Interested in mode shift and systemwide impacts (1)  Need to understand pedestrians crossing roads since this is when they are exposed (1)  Experience using many different travel modes and experience with the agency’s road system (1) Reasons given by interviewees for rating certain topics as low priorities were:  Lack of high pedestrian volumes (6)  Rely on other qualities (e.g., required crosswalk widths could vary depending on the environment) (1)  Research already exists and doesn’t need updating (e.g., LOS crossing a street and walking along a street, and there is a path width calculator) (1)  Prefer impedance to walking rather than crosswalk width analysis (1)  Very difficult to analyze systemwide connectivity (1) Other Thoughts and Interest in Future Project Participation Additional Thoughts Interviewees were asked whether there were any other issues related to pedestrian volume counting, pedestrian safety countermeasures, or pedestrian QOS that had not been covered in the interview that they would like to bring up. Comments included:  How much influence does a jurisdiction have for imposing development requirements—how far away from the site can they require pedestrian improvements?  Perception-based measures creates a discontinuity with other [operations-based] HCM methods.

247  How to create a robust counting program?  Safety countermeasure trade-offs of perceived and actual safety.  There’s not a lot of info out there on systematic standardized approach to a count program. There’s a lot of info on a count location, volumes at project location, but what’s best practice for how to measure across city, location selection, types, permanent count site, what to do with info.  It would be beneficial to receive additional guidance concerning the HCM methods and how to count pedestrians and groups of pedestrians.  Countdown timer lags between phases, LPIs – how beneficial are they for managing traffic? Crossing the street is much more important to evaluate than walking along a street. Look at pedestrian traffic in both directions, and also car traffic. Look at the pedestrian environment and potential improvements and impacts on car traffic. Pedestrian recall vs. pushbuttons; pushbuttons may lead to lower QOS and may need to evaluate it.  Need for HAWK signal & actuated ped signal methodology.  Regarding one-way streets, crossing a street with traffic coming at you from fewer directions might be better from both QOS and safety viewpoints.  Does the current HCM model need to be so complex?  Standardized intersection pedestrian count method. Currently only doing screenline counts.  Might be good to consider crowdsourced pedestrian volume sources; getting good pedestrian volume data at some point especially because of the unconstrained movements of pedestrians crowdsourcing might give us better information on pedestrian movements. So it would be good to include in this project if we can.  Pedestrians in rural areas is big concern for my agency. There are urban areas that act and look like rural areas, dotted all over the state. Stretches that have major rivers or highways that are barriers, not many crossing opportunities. Have to go a long way to get to a traffic light to cross. Desire lines are everywhere. People are crossing there. My agency has allowed little gaps in road barriers, slight break allows people to filter through it. No other signs or crosswalks. People are on their own.  We also have tribal communities which are even more rural, so trying to think about pedestrian networks in those communities are needed. Trying to address equity in priority populations.  Needs to be more study on how ped/bike count collection methods can be incorporated with existing infrastructure (traffic cameras, etc.)?  Heat maps to identify desire lines that don’t line up with existing sidewalks and pedestrian crossings. Future Project Participation All interviewees said that we could contact them with follow-up questions, if necessary, and all would like to receive a copy of the interview results, following panel approval of them. All but five were interested in participating in the peer exchange.