The extension of driving automation systems to transit buses has the potential to alter the urban transportation landscape, yet it can result in potential hazards that need further consideration by manufacturers and transit operators. Researchers applied key concepts from the voluntary functional safety standard, International Organization for Standardization (ISO) 26262, to identify potential vehicle-level hazards, assess their risks, and develop functional safety measures for driving automation applied to a generic 40-ft transit bus. Automation levels considered ranged from Level 0 (no driving automation) to Level 2 (partial driving automation) as defined by Society of Automotive Engineers (SAE) recommended practice J3016. This study assessed system interfaces with bus-specific systems, and considered the transit bus operating environment, which includes standing and unrestrained passengers, and a high number of vulnerable road users present. Analysts identified 18 potential vehicle-level hazards, to which the ISO 26262 risk assessment process was applied. Fifteen hazards were common to both light vehicles and transit bus driving automation systems, and three hazards were noted as being particularly relevant to transit buses:
- Vehicle motion when passenger door is open,
- Vehicle too far from curb at station / stop, and
- Excessive vehicle roll, due to transit buses’ higher center of gravity
- Establish guiding principles in advance to carefully consider tradeoffs such as maximum deceleration limits in protecting passenger and external vulnerable road users. System developers should define certain guiding principles that help to assess likelihood and maximum injury severity for both vulnerable road users and bus occupants and define deceleration parameters.
- Maintain acceleration and deceleration levels under permissible limits. Driving automation systems that control propulsion and braking should establish a limit on acceleration and deceleration in such a way that ensures safety of seated, unsecured standing passengers and displaced objects that could potentially cause any harm.
- Ensure the automation technology has real-time awareness of door status. Adaptive Cruise Control (ACC) and Traffic Jam Assist (TJA) technologies may have the potential to resume accelerating while passengers are still boarding or alighting the bus. The ACC and TJA algorithms should have real time awareness to restrict bus movement while passengers may be moving in or out of the doors.
- Update calibration parameters to reflect the physical characteristics of the bus if applying automation systems developed for other vehicles to transit buses. Consideration should be given to physical dimensions which may necessitate smaller margins for intervention in automated bus lane keeping and lane centering systems. Parameters should be calibrated to allow for adequate stopping distance and wider turns based on transit bus characteristics.
- Ensure that pedestrian detection systems have sufficient robustness and integrity in their installed locations on the bus. The detection algorithms should account for bus-specific sensor mounting locations to achieve a low false negative rate and high robustness in determining the correct braking distance.
- Evaluate sensor coverage to minimize obscured or obstructed areas in systems that rely on sensors. The physical dimensions of the bus may present challenges in placing sensors in a way that covers all needed views for each application’s optimal performance. Additional measures for each application may be needed to ensure that sensor coverage is sufficiently robust for safe application operation.