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.
8 Bus collision and fatality rates have followed a generally increasing trend, while bus injury rates have slightly decreased over the past decade, as shown in Figure 1 through Figure 3. Casualty and liability costs have also trended upward over the past decade, as shown in Figure 4. To improve safety, transit agencies have implemented various onboard technologies to provide additional information to operators and to mitigate collisions through automated responses. Technologies used to reduce accidents include, but are not limited to, the following: ⢠Driver vision systems ⢠Forward collision warning (FCW) ⢠Mitigation or automated emergency braking ⢠Lane departure warning (LDW) ⢠Electronic stability control (ESC) ⢠Pedestrian and other vulnerable road user detection ⢠Other forms of driver assistance systems Data trends substantiate the need to improve the safety of transit bus travel; the deployment of technologies, such as those listed previously, may play a role in safety improvements. This literature review explores the existing collision avoidance technologies used in transit bus appli cations, followed by examples of technologies that have potential applicability to transit buses but that are currently available only for other modes, such as personal vehicles or heavy trucks. Existing Bus Collision Avoidance Technologies This section of the literature review will introduce existing technologies that either are in use by transit agencies or have been developed for the transit market and are being piloted or other wise evaluated. Myriad technologies for bus collision avoidance have been developed for the transit market, including FCW, mitigated or automated emergency braking, LDW, electronic stability control, pedestrian and other vulnerable road user detection, and other forms of driver assistance systems, including videoÂbased driver monitoring systems. These technologies are designed to improve driver situational awareness and reaction time, with the ultimate outcome of reduced collision events and corresponding reductions in injuries, fatalities, and liability costs for the agency. Many of these technology initiatives and their application are elaborated on in the section on review of collision avoidance technologies, which discusses results of using these technologies. Collision Avoidance Warning Systems A study titled Active Safety-Collision Warning Pilot in Washington State introduces a system called the Rosco VQS4560 Mobileye Shield+, which uses Mobileye technology from Intel and C H A P T E R 2 Literature Review
Literature Review 9 Source: NTD S&S-40 data (NTD 2018). 0.0 0.4 0.8 1.2 1.6 2.0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 B us C ol lis io ns p er M ill io n R ev en ue M ile s Figure 1. Transit bus collision rate trend. Source: NTD S&S 40 data. 0 1 2 3 4 5 6 7 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 B us In ju rie s pe r M ill io n R ev en ue M ile s Figure 2. Transit bus injury rate trend. Source: NTD S&S-40 data (NTD 2018). 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 B us F at al iti es p er H un dr ed M ill io n R ev en ue M ile s Figure 3. Transit bus fatality rate trend.
10 Current Practices in the Use of Onboard Technologies to Avoid Transit Bus Incidents and Accidents Shield+ developed by Rosco (Lutin et al. 2017). This system, used for a pilot in Washington State, is a collision avoidance warning system (CAWS) that uses internal and external cameras to cover each identified blind zone for the driver. Its purpose is to âprovide coverage of blind zones where vulnerable road users may be hidden from the driverâs view and by alerting the driver to avoid potential collisionsâ (Lutin et al. 2017). When a pedestrian, bicyclist, vehicle, or other obstruction is identified, the system is designed to alert and draw the attention of the driver to the potential collision. The authors report that the technology changed driversâ performance and significantly reduced alerts. While the system has the ability to detect lane departures, the agencies determined that it was better to disable the LDW because buses frequently change lanes as part of the route system. The Abellio London pilot using Mobileye from Intel kept the LDW system activated, contributing to a 29 percent reduction in avoidable collisions (SmartCitiesWorld 2018). The London pilot included the use of forward collision warning, pedestrian and cyclist collision warning, headway monitoring and warning, lane departure warning, and speed limit indica tor and traffic sign recognition. SixtyÂsix buses, operating on three routes, were equipped with Mobileye. In addition to the 29 percent reduction in avoidable collisions, there was also an estimated 60 percent reduction in injuries that may have resulted without the technology (SmartCitiesWorld 2018). A paper titled âInÂVehicle Stereo Vision System for Identification of Traffic Conflicts Between Bus and Pedestrianâ reports on a study that used stereoscopy (a visual technique that produces threeÂdimensional imagery) to âobtain a depth map related to the scene in front of the vehicleâ (Cafiso, Di Graziano, and Pappalardo 2016). Researchers used a TYZX DeepSea G3 Embedded Vision System to create a stereoscopic image that detects conflicts. When a conflict is detected, â[t]he target point is then automatically tracked in the successive frames in order to obtain mea surements covering the entire traffic conflict eventâ (Cafiso, Di Graziano, and Pappalardo 2016). The paperâs authors indicate that this system has the capability to activate an inÂvehicle system warning and driving assistance in real time, much like what is seen with VQS4560 Mobileye Shield+ and Intel Mobileye. Video-Based Driver Risk Management Systems In their paper titled Evaluating the Effectiveness of VideoÂBased Driver Risk Management Systems on Transit Safety, Michael Litschi and Peter Haas (2012) review the use of video recorders on transit vehicles, particularly the DriveCam system. These video recorders provide Source: NTD Annual Operating Expense data (NTD 2017). $- $50 $100 $150 $200 $250 $300 $350 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 C as ua lty a nd L ia bi lit y C os ts p er T ho us an d R ev en ue M ile s Figure 4. Transit bus casualty and liability cost trend.
Literature Review 11 transit agencies with moving imageâbased evidence for any type of event deemed unusual. The study finds that participating drivers at the two firms âreduced the mean frequency of recorded safetyÂrelated events per 10,000 vehicle miles traveled by 37 percent and 52.2 percentâ (Litschi and Haas 2012). The study also describes successful uses of videoÂbased recordings, including using the videos as training tools, incentivizing safety awareness using captured videos, and encouraging selfÂevaluation and driver observation. Hickman and Soccolich (2014) had similar results in their study of the DriveCam system for Virginia Tech University. They find that âtransit operators are changing their behavior because of the DriveCam system and learning to avoid the risky driving behaviors that cause an event to be captured and scoredâ (Hickman and Soccolich 2014). The Hickman and Soccolich study reinforces the results of Litschi and Haas: it finds 35.8 percent, 35.3 percent, and 34.7 percent reductions in heavy truck and bus injury crashes in 2010, 2011, and 2012, respectively (Hickman and Soccolich 2014). Review of Collision Avoidance Technologies Piloted or Deployed on Transit Buses The previous section explored existing technologies to prevent and mitigate collisions and discusses their functionality. This part of the literature review examines several of the previously described technologies that transit agencies throughout the United States have implemented or piloted. The technologies reviewed here are driver assist, connected vehicles, Rosco Mobileye Shield+, DriveCam, and Smart Drive. Their application and effectiveness are discussed, as well as any issues identified by the piloting or deploying agency, or in other research findings. Minnesota Valley Transit Authority Driver Assist System In the report titled Evaluation of Automated Vehicle Technology for Transit, CUTR (2015) describes a driver assist system that was developed at the University of Minnesota and deployed by the Minnesota Valley Transit Authority (MVTA). Minnesota state law allows buses âto use highway shoulder lanes when speeds in the general purpose lanes drop below 35 miles per hour.â The purpose of the system is to âencourage the bus drivers to use the shoulder lanes during inclement weather when the shoulder boundaries are obscured by snowâ (CUTR 2015). Accord ing to CUTR, âThe bus driver receives three types of feedback: visual, tactile, and haptic.â Using the provided headsÂup display and visual mirror, the driver assist system alerts bus operators âto any vehicles getting too closeâ and helps âthe drivers merge from the shoulder lane back into the general purpose lanesâ (CUTR 2015). Haptic feedback through the seat and steering provide tactile feedback for lane departure while the bus operator is driving on the shoulder. Upon review of the implemented system, MVTA found that âthe bus drivers stayed in the shoulder 4.3 percent longer, drove 3.5 miles per hour faster, and reduced their side to side movement by 4.7 inchesâ (CUTR 2015). U.S. DOT Connected Vehicle Pilot Model Deployment The Transit Safety Retrofit Package Development, Final Report discusses a U.S. DOT pilot deployment of connected vehicles (Zimmer et al. 2014). The pilot equipped light/mediumÂduty vehicles, trucks, and transit buses with connected vehicle technology. Using vehicleÂtoÂvehicle technologyâwhich includes FCW, emergency electronic brake lights, curve speed warning (CSW), pedestrian in signalized crosswalk warning (PCW), and vehicle turning right in front of bus warningâand vehicle to infrastructure technologyâwhich includes PCWâthe pilot program concluded with lessons learned for any future study or activity with connected vehicles. The lessons learned indicate that PCW technology is unable to distinguish between pedestrians
12 Current Practices in the Use of Onboard Technologies to Avoid Transit Bus Incidents and Accidents and vehicles, should not be dependent upon being on specific routes, and should be suppressed after the bus enters the crosswalk. The authors are also critical of using GPS for PCW and vehicle rightÂturn warnings, citing limitations of GPS technology, especially in downtown urban settings. In addition to the vehicle rightÂturn warning, the authors note that only the brake pedal was used in the application of this technology, and that gear position should be taken into account as a basis of a driverâs intent to proceed. As a result of these findings, Zimmer et al. report a low rate of driver acceptance of the technology. They emphasize that âfirmware and software reliability is critical to system performance,â and they note that improvements must be made to the display for power and brightness and that adjustments to the amount and type of data collected are needed to improve effectiveness (Zimmer et al. 2014). As part of a redeployment by the U.S. DOT, Zimmer et al. (2014) indicate that changes included verbal notifications rather than beeps, longer warning signals on the inÂvehicle display, adjustment of the device position to bring it closer to the driver, and mitigations to resolve power issues. They believe these improvements may help with driver acceptance. For the PCW, improvements related to the accuracy of distinguishing a pedestrian from a vehicle, knowing when the transit vehicle has entered the crosswalk area, seeing the vehicleâs position relative to a crosswalk, and modifying the alerts for when the driver has passed the crosswalk (Zimmer et al. 2014). Finally, for the vehicle rightÂturn warning, the settings were adjusted to ensure that alerts would not occur unless the vehicle is in forward drive. Texas A&M University, Rosco Mobileye Shield+ As part of a study on automated and connected vehicles for the Texas Department of Transportation and the FHWA, a pilot using the Rosco Mobileye Shield+ was conducted on the Texas A&M University (TAMU) campus over a 27Âday period in 2015 and 2016. The pilot operated from 7:00 a.m. to 5:00 p.m. on a route that âtraverses several crowded areas on the TAMU campusâ (Odell and Turnbull 2017). âThe preliminary assessment of the Mobileye Shield+ found the videoÂbased collisionÂwarning system accurately detected pedestrians and bicyclists in close proximity to the bus during the 27Âday periodâ (Odell and Turnbull 2017). There were no false alarms over the course of the pilot, and video collected was âusable and viewable on 37 of the 41 eventsâ detected (Odell and Turnbull 2017). The study states that bus operations personnel had an overall positive reaction to the system; however, there were concerns âon whether or not alerts would give enough time for driver to react. . . . Drivers also stated the need for a similar system that would work at nightâ (Odell and Turnbull 2017). With regard to the latter concern, the study notes that Mobileye Shield+ is not advertised as functional in lowÂlight environments. Since then, the technology has evolved and there has been successful field testing of Mobileyeâs latest system, specifically by the Virginia Tech Transportation Institute for Pierce Transit. Washington State, Rosco Mobileye Shield+ In addition to its pilot at Texas A&M University, the Rosco Mobileye Shield+ system was also piloted in Washington State, as noted in the previous section. A TRB study titled Active Safety-Collision Warning Pilot in Washington State discusses the installation by the Washington State Transit Insurance Pool (WSTIP) of the Rosco Mobileye Shield+ system on 38 transit vehicles at WSTIP member agencies over the course of three months in 2016 (Lutin et al. 2017). Of the 38 equipped buses, a control group of an undefined number was set to operate in âstealth mode,â meaning the vehicles were fully equipped and operating with the Mobileye Shield+ system, but no alerts were given to the drivers. In addition to examining how well the system operated, the study included surveys distributed to the bus operators, tests for false negatives
Literature Review 13 and false positives, measurement of the performance of collision avoidance systems, analysis of comparisons with historical claims, and a costÂeffectiveness analysis of a full implementation. Over the course of the pilot, âBuses equipped with Shield+ systems logged 352,129 miles and 23,798 operating hours during the official pilot data collection periodâ and âNo Shield+ equipped buses were involved in any collisions with bicyclists or pedestriansâ (Lutin et al. 2017). The pilot study concludes that there was a âlarge reduction in nearÂmiss events for CAWSÂequipped busesâ (Lutin et al. 2017). The following results were found in each of the testing areas. According to driver surveys, most drivers (63 percent) found the system to be distracting, but an even larger percentage (67 percent) said they would rather have this technology while driving (Lutin et al. 2017). Concerning false positives, there were reports of a pedestrian collision warning being âgenerated by movement of the bus towards an object similar in shape to a pedestrianâ and when pedestrians and bicyclists were âmoving parallel to and on the left of the bus in either the same or oppo site directionâ (Lutin et al. 2017). Some false negatives were reported, but âno strong patterns emergedâ (Lutin et al. 2017). Overall, the study of the Mobileye Shield+ found that vehicles with the system installed âexperienced 71.55 percent fewer collision warningsâ and resulted in 43.32 percent fewer combined pedestrian collision warning and pedestrian detection zone warnings (Lutin et al. 2017). Lutin et al. (2017) hypothesize that the CAWSÂequipped buses made drivers more sensitive to conditions that triggered warnings, and those drivers were able to anticipate those conditions and avoid triggering the CAWS indicators. Thus, Lutin et al. deduce that CAWS might be able to reduce collisions by increasing driver awareness of potential conditions that might lead to a crash. The study estimates that the annual net benefits from collision claims reduction for WSTIP members would increase from $1,099,262 in year 5 to $2,102,473 in year 14 (Lutin et al. 2017). For the lower bound, a loss, â$4,232, was estimated in year 5, but benefits became positive in year 6, and increased to $998,979 by year 14 (Lutin et al. 2017). Onboard Camera Applications for Buses TCRP Synthesis 123: Onboard Camera Applications for Buses states there are between seven and 10 cameras on 87 percent of the Southeastern Pennsylvania Transportation Authorityâs (SEPTA) transit fleet (1,245 vehicles) (Thomson, Matos, and Previdi 2016). As a result of its implementation of onboard cameras, âthe authorityâs claims and litigations have been on the decline and have saved SEPTA more than $40 million annuallyâ (Thomson, Matos, and Previdi 2016). The video system provides SEPTA the ability to âreview all incidents, request and retrieve video for review and analysis, and reconcile reports and video, creating a file typically within 24 hours of the incident and long before most claims can even materializeâ (Thomson, Matos, and Previdi 2016). Thomson, Matos, and Previdi (2016) note that â90 percent of the videos are used for legal purposes, 9 percent assist the Transit and City of Philadelphia Police, and 1 percent support customer service and operations compliance.â Litschi and Haas (2012) identify Capital Metro, New Jersey Transit, Pace Suburban Bus Service, San Francisco Municipal Transportation Agency, Washington Metropolitan Area Transit Authority, First Transit, MV Transportation, and Veolia as users of the DriveCam camera system and the Los Angeles County Metropolitan Transportation Authority (Metro) as a user of the SmartDrive Camera System. Hickman and Soccolich (2014) in their report for Virginia Tech describe DriveCam as âa closedÂloop behavior modification system with several steps to assure positive outcomes.â Both reports reflect the same sentiment about DriveCam and SmartDrive; Litschi and Haas (2012) describe the use of an onboard camera system as a training
14 Current Practices in the Use of Onboard Technologies to Avoid Transit Bus Incidents and Accidents tool to identify risky driving behaviors of the bus operators and âto improve operational safety and reduce traffic accidents by increasing operatorsâ awareness of their unsafe driving habits and coaching them to modify these unsafe driving behaviors.â As a result, each transit authority saw a decrease in the number of collisions per million revenue miles, passenger injuries per million revenue miles, preventable accidents, traffic violations, risky behaviors, and collisions. Specific details from each agency can be gleaned from the full report by Haas and Litschi. Litschi and Haas (2012) also report that Pace Suburban Bus was able to use the DriveCam system to âcombat fraudulent claims from passengers and operators and help dismiss traffic tickets.â It is important to note that, while Capital Metro initially saw a 50 percent drop in collisions and passenger injuries after the systemâs implementation, these rates returned to their previous levels or higher in the following years (Litschi and Haas 2012). The report did not disclose possible reasons for the increase to previous levels. Metro (Los Angeles), in contrast, has seen a â30 percent reduction in events with safety concerns . . . but they have not seen a consistent reduction in accidents that can be attributed to implementation of SmartDriveâ (Litschi and Haas 2012). This section reviewed transit agency experiences in piloting technologies and evaluating their ability to reduce the frequency and severity of transit bus collisions. The final section of Chapter 5 describes the literature related to collision avoidance technologies that have been developed and implemented for the personal and commercial vehicle market.