Current Practices in the Use of Onboard Technologies to Avoid Transit Bus Incidents and Accidents (2020)

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.

15 Industry Survey Findings For this synthesis, a survey was conducted of select transit agencies to obtain their perspective and experiences regarding onboard collision avoidance technologies aimed at reducing bus accidents and incidents. The 38-question survey questionnaire (included in Appendix A) was sent to 55 different transit agencies throughout the United States. The transit agencies were chosen on the basis of previous recognition of their pilot demonstrations, as well as known technology implementations within their systems aimed at reducing transit collisions. They were chosen also on the basis of size and geographic distribution, previous interest in the topic of transit collision reduction, and reliable contacts within the agencies known by the research team. This method ensured that there would be both regional and size variation in systems represented from across the nation, as well as an acceptable survey response rate. Forty-four of the transit agencies responded, yielding an 80 percent response rate. The survey was categorized into the following four sections: 1. System Characteristics 2. Accident Review and Safety Risk Assessment 3. Use and Acceptance of Technologies 4. Agency Model Practices or Other Experiences Survey Data Summary A summary of survey responses by section is presented here. System Characteristics This first section requested respondents to provide contact information as well as general system characteristics, such as the number of operators by mode and full- or part-time status, and the number of vehicles in their fleet by size of vehicle. The transit agencies that responded were geographically dispersed across 23 states, as reflected in Table 1. The survey respondents also represent a wide range of agency sizes, from two full-time bus operators with Skyline Bus in Montana to 12,217 full-time bus operators with New York City Transit. Bus fleet sizes also vary from fewer than 10 total buses in operation to over 5,700 buses in operation. Accident Review and Safety Risk Assessment The accident review and safety risk assessment section of the survey inquired about the examination and subsequent tracking of causal or contributing factors in safety event C H A P T E R 3 Survey Results

16 Current Practices in the Use of Onboard Technologies to Avoid Transit Bus Incidents and Accidents investigations. As shown in Figure 5, 88 percent of the transit agencies that responded indicated that they do track causal or contributing factors. When asked if their review of incidents includes an examination of technology systems to determine if the technology was working as intended, 81 percent of transit agency respondents indicated that they do examine the technology, as displayed in Figure 6. Transit agency representatives were asked their opinion on the most prevalent causal factors of preventable safety incidents within their agency. Human factors, specifically related to not following policy or procedure, were the most prevalent causal factor, at 93 percent of the transit agencies that responded. Agencies were asked to select all causal factors that apply, so the sum California Alameda–Contra Costa Transit District (AC Transit) Los Angeles County Metropolitan Transportation Authority (LAMetro) San Francisco Municipal Transportation Agency (SFMTA) Santa Clara Valley Transportation Authority (VTA) Colorado Greeley-Evans Transit Connecticut Greater Bridgeport Transit (GBT) Norwalk Transit District District of Columbia Washington Metropolitan Area Transit Authority (WMATA) Florida Central Florida Regional Transportation Authority (LYNX) Hillsborough Area Regional Transit Authority (HART) Miami-Dade Department of Transportation and Public Works (Miami-Dade Transit) Jacksonville Transportation Authority (JTA) Lee County Transit (LeeTran) Georgia Metropolitan Atlanta Rapid Transit Authority (MARTA) Indiana Indianapolis Public Transportation Corporation (IndyGo) Iowa Ames Transit Agency (CY Ride) Maryland Maryland Transit Administration (MTA) Massachusetts Massachusetts Bay Transportation Authority (MBTA) Minnesota Metro Transit Missouri City Utilities of Springfield Transit (CU Transit) Montana Fort Peck Transit Skyline Bus North Carolina GoTriangle Greensboro Transit Authority (GTA)—Operated by Keolis New Jersey New Jersey Transit (NJT) New York New York City Transit (NYCT) Ohio Greater Cleveland Regional Transit Authority (RTA) Southwest Ohio Regional Transit Authority (SORTA) Toledo Area Regional Transit Authority (TARTA) Oregon Lane Transit District (LTD) Tri-County Metropolitan Transportation District of Oregon (TriMet) Pennsylvania Port Authority of Allegheny County Southeastern Pennsylvania Transportation Authority (SEPTA) South Carolina Central Midlands Regional Transit Authority (CMRTA) South Dakota River Cities Public Transit Texas Dallas Area Rapid Transit (DART) Metropolitan Transit Authority of Harris County (METRO) Trinity Metro Utah Utah Transit Authority (UTA) Washington Community Transit (CommTrans) King County Metro Transit (KC Metro) Pierce County Public Transportation Benefit Area Corporation (Pierce Transit) West Virginia Mid-Ohio Valley Transit Authority (MOVTA) Wisconsin City of Madison Metro Transit (Madison Metro) Table 1. Survey participants. 88% 12% Yes No Figure 5. Transit agencies that examine causal or contributing factors in safety events. 81% 19% Yes No Figure 6. Transit agencies that examine technology postevent to determine proper functionality.

Survey Results 17 of the responses is greater than 100 percent, as shown in Figure 7. Some of the responses that were detailed as “other” included operator stress, schedule pressures, equipment failure, complacency, blind spots in cutaways, and weather conditions. The last question in the “Accident Review and Safety Risk Assessment” section of the survey asked how transit agencies identify safety risks. Respondents were asked to select all mechanisms that apply. Figure 8 shows that more than 75 percent of all responding agencies track safety risks through internal safety reviews, customer reports or complaints, tracking data trends, employee reporting (verbal, comment cards, or general observations), and accident review reports. Fifty percent of the responding agencies use employee close-call or near-miss reporting to identify safety risks. Sixteen percent of transit agencies specified some “other” way that they identify safety risks, which included safety committees, the “See & Say App,” and reaching out directly to operators to gather safety hazard data. Use and Acceptance of Technologies Use of Technologies. Once the safety-related issues are identified, the transit agencies use combinations of several methods to mitigate those issues. Transit agency representatives were Fatigue Other (please describe) Human factors – training related Disobeying/not observing traffic laws Distractions Human factors – not following policy or procedure 0% 20% 40% 60% 80% 100% Figure 7. Most prevalent causal factors of preventable safety incidents. 0% 20% 40% 60% 80% 100% Other (please specify) Review of bus operator reporting software/telemetric systems (e.g., DriveCam/LYTX or SmartDrive) Employee close call or near miss reporting Internal safety reviews Tracking data trends Customer reports/complaints Other employee reporting (verbal, comment cards, supervisor or dispatch observations) Accident review reports Figure 8. How transit agencies identify safety risks.

18 Current Practices in the Use of Onboard Technologies to Avoid Transit Bus Incidents and Accidents asked to identify which safety technologies are utilized, or have been considered, by their agency. Figure 9 shows that onboard security cameras are the most considered technology, with 93 per- cent of agencies indicating they have at least considered them. About 14 percent of agencies reported using or considering a modified brake interlock system. Further refining the questions to focus on onboard collision avoidance technologies, the survey asked transit agencies to identify the specific collision avoidance technologies that they have implemented. Agencies were asked to select all collision avoidance technologies that apply; therefore, transit agencies were asked to select “N/A” in order for the survey to capture the share of respondents that have not implemented any collision avoidance technologies. Survey responses are displayed in Figure 10. Passive technologies, such as exterior vehicle security cameras and rear vehicle treatments were the most common, while active technologies such as automated braking were much less common. About one in three of the transit agencies that responded to the survey indicated that it has not implemented any collision avoidance technologies. Transit agencies were asked what collision avoidance technologies had been considered but not utilized at their agency. If no technologies fell into this category, transit agencies were prompted to skip the question. Twenty responses were received, but one of the responses was 0% 20% 40% 60% 80% 100% Modified brake interlock Other (please specify) Door speed interlock Electronic data recorders Driver monitoring/reporting programs Front door brake interlock Video data recorders Stop announcements Vehicle tracking system Onboard security cameras Figure 9. Safety technologies in use (or considered by) transit agencies. 0% 10% 20% 30% 40% 50% 60% Automated emergency braking Lane departure warning (LDW) Electronic stability control (ESC) Forward collision warning (FCW) Pedestrian and other vulnerable road user detection Driver-view external video monitoring/detection Other (please specify) Pedestrian warning devices Special rear vehicle treatments N/A Exterior vehicle security cameras Figure 10. Collision avoidance technologies implemented.

Survey Results 19 more of a clarification as to how rural transit agencies rely upon their state DOT to procure the vehicles they use. The other 19 responses primarily identified the use of passive collision avoidance technologies, such as lane departure warnings, pedestrian detection, audible turning warnings, and driver fatigue detection technologies. Two responses indicated the consideration of an active collision avoidance system autonomous emergency braking; however, the respon- dents indicated that cost, inflexible contracting agreements, and false positive alerts were the main reasons for not implementing the technology. Procurement and Installation of Technologies. Concerning the procurement and installa- tion of collision avoidance technology, transit agencies were asked if the technology was offered by the transit vehicle manufacturer and installed during final vehicle assembly, prewired by the transit vehicle manufacturer during final assembly, installed “after market” by the technology vendor, or installed on modified transit vehicles at the transit agency by either vendor technicians, agency personnel, or both. As shown in Figure 11, the most common (46 percent) installation process occurred at the transit agency and was performed by agency personnel, vendor techni- cians, or both. Nearly 30 percent of the respondent agencies indicated that the technology was offered by the transit vehicle manufacturer and installed during final vehicle assembly. Four percent of respondents indicated that the vehicle manufacturers prewired the transit vehicles during final vehicle assembly. Agencies were asked how long each collision avoidance technology has been in use at their agency, and the responses ranged from “started this week” to 9 years, with an average time of 2.7 years. Additionally, to gain an understanding of the operating environment of the buses equipped with collision avoidance technology, respondents were asked if the equipped buses operate in urban, suburban, or rural environments. If the buses operate in more than one type of environment, respondents were asked to select all that apply. Respondents indicated that over 85 percent of the buses equipped with collision avoidance technology operate in an urban environment, while just under 29 percent operate in a rural environment, as displayed in Figure 12. Less than half of the agencies that responded, 47 percent, indicated that the collision avoid- ance technologies implemented at their agency resulted in documented safety improvements. Some of the respondents that indicated no documented safety improvements commented that it was too early to determine the results of the technology. Several agencies reported decreased collision rates, while other agencies were more specific in the decreases by the type of collision, such as a left-turn pedestrian collision. As a follow-up, agencies were asked how safety improvements are measured within their agency. Nearly 70 percent of respondents indicated that safety improvements are measured using specific performance metrics with values, as shown in Figure 13. 46% 29% 21% 4% The transit vehicles were modified at the agency location by vendor technicians, agency personnel, or both. The technology was offered by the transit vehicle manufacturer and installed during final vehicle assembly. The transit vehicles were provided to the technology vendor and application was installed “after-market.” The transit vehicle manufacturer pre-wired vehicles during final vehicle assembly. Figure 11. Procurement and installation process.

20 Current Practices in the Use of Onboard Technologies to Avoid Transit Bus Incidents and Accidents Transit agencies were asked why they considered the implementation of the collision avoid- ance technology, and they were specifically asked if the consideration was in response to specific safety events or trends. Responses varied from proactive consideration due to benchmarking with peer agencies and additional funding available to reactive consideration due to increased trends in pedestrian collisions. Other reasons for considering the technology included the Vision Zero Initiative, driver requests, and “basic logic.” When asked about barriers to implementation, 38 percent of respondent agencies indicated that they faced no barriers to implementation. The other 62 percent of agencies that did experi- ence some type of barrier reported cost, retrofitting challenges, union and operator resistance, nonworking equipment, and limited market competition as the barriers faced, as shown in Figure 14. It is worth noting that one respondent agency that stated it faced no barriers to implementation did indicate that the proactive inclusion of union and labor representatives in the technology selection process led to ownership and approval of the technology once the implementation occurred. Sixty-five percent of survey respondents indicated that they included labor representatives in the discussion and decision-making process associated with the demonstration or deployment of collision avoidance technologies. 85.7% 52.4% 28.6% 0% 20% 40% 60% 80% 100% Urban Suburban Rural Figure 12. Operating environments of buses equipped with collision avoidance technology. 0% 20% 40% 60% 80% Research projects Other (please specify) Anecdotal evidence Perception Specific performance metrics with values Figure 13. How safety improvements are measured.

Survey Results 21 The majority of respondent agencies (85 percent) reported that they piloted/tested the collision avoidance technology prior to full deployment on their bus system. Sixty-three percent of respondents considered different vendors of the same type of collision avoidance technology prior to the technology deployment at their agency. The responses to previous survey questions hint that the share could have been higher if there were more market competition. Of the respondent agencies that have implemented an onboard collision avoidance tech- nology, 53 percent indicated that they strategized implementation either by the route or type of operating environment, or by bus type or age. The routes and operating environments with higher potential for interaction were more common strategic factors for onboard collision avoidance technology installation than bus type, as shown in Figure 15. Acceptance of Technologies. Of the respondent agencies that had implemented some type of collision avoidance technology, 95 percent reported that they provided technology training or orientation to employees. The descriptions of the training included training for both installation and operation, with delivery methods that range from educational handouts to classroom and hands-on training. No Barriers 38% Cost/ROI Concerns 19% Retrofitting Challenges 14% Union/Operator Resistance 14% Non-working Equipment 10% Limited Market Competition 5% 62% Figure 14. Barriers to Implementation (ROI = return on investment). No 47% Yes - by Route/Operating Environment 79% Yes - by Bus 21% Yes 53% Figure 15. Strategized installation.

22 Current Practices in the Use of Onboard Technologies to Avoid Transit Bus Incidents and Accidents Fifty-five percent of the agencies that piloted or implemented a collision avoidance tech- nology surveyed their operators to obtain feedback on the technology. Figure 16 displays the general survey results. Of the agencies surveyed, 36 percent reacted favorably to the collision avoidance technology, 29 percent had mixed reactions, 21 percent received no feedback, and 14 percent reacted negatively. While 37 percent of respondent agencies reported noticeable changes in operator or other employee reactions to the collision avoidance technology over time, several agencies commented that sufficient time had not passed to make that determination. Input from the operators led to warning volume adjustments and solutions to light detection and ranging (lidar) challenges. When asked if the deployment of any technologies had resulted in cost savings to the agency through reduced liability claim payouts or decreased incidents, 44 percent of the respondents indicated a cost savings. However, many of those that did not report a cost savings have either not yet calculated the costs or have not had sufficient test time to make the determination. More than half, 53 percent, of respondent agencies have considered the pros and cons of competing technologies. Vendor Experience. When asked if their vendor agreement included terms for trouble- shooting, maintenance, or technical support, 86 percent of the respondent agencies indicated their agreement did include those terms. Some agreements included temporary support or oversight, while others were described as warranties. Additionally, 74 percent of the respon- dents indicated that their vendor agreement did not include terms for upgrading technology as upgrades become available. This potential for upgrading is an important element because these technologies are evolving at a rapid pace. Agencies were asked whether the vendor removed all equipment if, following the initial deployment, the agency elected not to use the technology. Responses were split, with 47 percent indicating that the vendor removed all equipment and the other 53 percent reporting that the equipment was removed in-house. The majority of agencies, 89 percent, indicated that they did not do any community outreach related to the implementation of the technology. Agency Model Practices or Other Experiences As a final question, agencies were asked to provide any model practices or experiences, for both technology deployments and general agency best practices. The majority of responses were categorized into five main categories: pedestrian detection/alert technology, tracking and Not surveyed 45% Positive 36% Mixed Reaction 29% No Response/No Interactions 21% Negative 14% Surveyed 55% Figure 16. Employee survey results.

Survey Results 23 trending data, using technology to improve training, other technologies or safety improvements, and technology dissatisfaction. Pedestrian Detection/Alert Technology. Nearly 16 percent of the respondent agencies that offered some type of best practice as a response to the final survey question indicated that pedestrian turn warning or pedestrian alert systems were beneficial technologies that were implemented on their bus fleet. One respondent clarified that the pedestrian detection technology was implemented proactively on the agency’s newly procured electric buses because the quieter operation of the electric buses was a cause for concern to the agency. Tracking and Trending Data. The practice of tracking and trending data to guide the implementation of safety mitigation measures was defined as beneficial to 37 percent of the respondent agencies that provided some sort of model practice. One agency specifically uses hot-spot mapping to identify specific areas of concern to allow for focused intervention methods such as ride checks, radar checks, and coaching opportunities. Some of the respondents mentioned the specific benefits of driver monitoring technology that activates automatically upon changes in forces due to evasive maneuvering or a collision event. Survey respondents highlighted the value of intervening with targeted refresher training using the data collected from the driver monitoring programs. Additionally, the video collected from these driver monitoring technologies has been beneficial in complaint investigation and has contributed to decreased liability payouts. Using Technology to Improve Training. While many of the respondent agencies indicated the value of investing in collision avoidance and other types of technologies to improve safety, many (37 percent) also emphasized the importance of training. Respondents emphasized that training is important in all stages of employment, from new employee training to refresher training. Industry Survey Conclusions The summarized survey responses are representative of 44 various-sized transit agencies from 25 U.S. states and the District of Columbia. These transit agencies have demonstrated improved safety cultures within their agencies. The majority of responding agencies, 88 percent, track causal or contributing factors in safety events, and 81 percent of agencies check the technology that is in place to determine if the technology was working properly at the time of the event. The most common causal factor, in the opinion of the respondents, was human factors related specifically to not following policies or procedures. When asked how risk is identified within their agencies, the majority of respondents reported multiple methods, such as reviewing accident reports, employee reporting, tracking performance metrics trends, customer complaints, and internal safety reviews. Fifty percent of the responding agencies indicated that they identify risk through employee close-call or near-miss reporting. The most commonly reported technology application on transit buses is the use of onboard security cameras, which are reportedly used by 93 percent of the responding agencies. Other common technologies include vehicle tracking systems, stop announcements, video data recorders, and front door brake interlocks. When asked specifically about collision avoidance technologies installed on their transit buses, half of the responding agencies indicated using exterior vehicle security cameras, 30 percent have implemented some sort of rear bus treatment to mitigate rear-end collisions, and 23 percent have implemented pedestrian warning technol- ogy on their buses. Thirty-two percent of the responding agencies have not implemented any type of collision avoidance technology on their transit buses.

24 Current Practices in the Use of Onboard Technologies to Avoid Transit Bus Incidents and Accidents For the agencies that responded that they did procure some type of collision avoidance technology, 46 percent indicated that the transit vehicles were modified at the transit agency by vendor technicians, agency personnel, or both, while 4 percent indicated that the manufacturer prewired the transit bus during final vehicle assembly. As expected, these collision avoidance technologies are typically installed on buses that operate in urban environments, where more transit buses operate and there is higher potential for interaction with other road users. The biggest barriers to implementation of collision avoidance technologies on transit buses are cost and return on investment concerns, retrofitting challenges, and union or operator resistance. Of the 55 percent of respondent agencies that surveyed their employees for feedback on the collision avoidance technology, 36 percent had mostly positive feedback, 29 percent had mixed reactions, and 14 percent had mostly negative feedback. In terms of vendor agreements, 86 percent of the agreements included terms for trouble- shooting, maintenance, or technical support, while 74 percent did not include any terms for upgrading the technology. As suggested earlier, this element is important, as these technologies are evolving at a rapid pace. Labor Survey Findings In addition to surveying safety representatives from 55 transit agencies, the synthesis team also developed and conducted a brief survey that the Amalgamated Transit Union (ATU) dis- tributed to its operator members through local union presidents. Of the 153 survey responses received, 30 respondents indicated they had driven a bus with some form of collision avoidance technology, and 25 of the 30 respondents indicated a willingness to share their experiences associated with each technology application. Survey question 2 provided a list of nine technologies and “other” and asked respondents to select all technology applications that were on buses they had driven. Therefore, the totals shown in the resultant Figure 17 exceed 100 percent when summed. It is important to consider the values that are associated with the figure shown, as the largest group of respondents—those who drove a bus with a camera-based pedestrian, bicycle, or vehicle detection technology— represents 17 individuals. Alternatively, one respondent drove a bus with a pillar camera, so any opinions expressed related to that technology are the opinions of one individual. The “other” 1 2 2 2 3 5 7 8 10 17 0% 20% 40% 60% 80% Pillar cameras Drive on shoulder lane keeping Connected vehicle Other (please specify) 360-degree cameras Audible pedestrian warnings Backup camera Forward collision avoidance Lane departure warning Camera-based pedestrian, bicycle, or vehicle detection Figure 17. Type of collision avoidance technology exposure.

Survey Results 25 types of technologies that were included in the survey responses included antilock brakes and a rear sensor on a paratransit vehicle that made an audible alert when an object was detected to be close to the vehicle. Once the ATU operator survey respondents identified the type of bus collision avoidance technology exposure, the next question asked respondents to indicate their experience with each of the technologies identified. The opinions of the respondents displayed in Figure 18 and Figure 19 are the shares of the number of respondents for each technology displayed in Figure 17. Survey questions 4 and 5 asked respondents to report if training had been provided on the technologies and, if so, their view of whether or not the training was sufficient. Respondents indicated that training on the technologies was provided 28 percent of the time. The technolo- gies with the highest share of training provided were those associated with connected vehicle 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Camera-based pedestrian, bicycle, or vehicle detection Lane departure warning 360-degree cameras Forward collision avoidance Backup camera Connected vehicle Rear sensor on paratransit bus Instrumental in avoiding events Helpful Figure 18. Positive technology feedback. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Forward collision avoidance Camera-based pedestrian, bicycle, or vehicle detection Lane departure warning Connected vehicle Pillar cameras 360-degree cameras Drive on shoulder lane keeping Audible pedestrian warnings Problematic due to false alarms Distracting Figure 19. Negative technology feedback.

26 Current Practices in the Use of Onboard Technologies to Avoid Transit Bus Incidents and Accidents applications. For those respondents who did receive training, 87 percent felt that the training they received was adequate, as shown in Figure 20. Finally, Question 7 of the survey asked respondents to rank their overall acceptance of each technology. Figure 21 illustrates the responses, which show generally low acceptance or favor- ability for each technology, with the exception of rear-collision sensors on the rear of paratransit vehicles. No Training 72% Adequate Training 87% Not Adequate 13% Training Provided 28% Figure 20. Technology training. R ea r se ns or on pa ra tr an si t bu s C on ne ct ed ve hi cl e B ac ku p ca m er a La ne de pa rt ur e w ar ni ng F or w ar d co lli si on av oi da nc e C am er a- ba se d pe de st ria n, bi cy cl e, or v eh ic le de te ct io n A ud ib le pe de st ria n w ar ni ng s P ill ar c am er a s 36 0- de gr ee ca m er as D riv e on sh ou ld er la ne ke ep in g U nf av or ab le to F av or ab le Figure 21. Overall technology acceptance.