Deploy side object detection systems for transit buses that have proven effectiveness in transit operating environments and been accepted by transit operators.
Experiences from the installation of Side Object Detection Systems on transit buses in Ohio, Utah and Washington, D.C.
Made Public Date
02/15/2010

69

Washington
District of Columbia
United States

109

Salt Lake City
Utah
United States

1028

Cleveland
Ohio
United States
TwitterLinkedInFacebook
Identifier
2010-00509

Side Object Detection System Evaluation: Final Evaluation Report

Background

Side object detection systems (SODS) have the potential to improve the safety of transit buses by helping transit operators avoid collisions with objects on the side of the bus. Data from the Federal Transit Administration (FTA) showed that almost half of all transit bus collisions occur on the side, costing an average of $3,660 in property damages per side collision. The transit operating environment presents challenges for a collision detection system because buses range into and out of proximity to side objects of various types, from signs, other vehicles, poles, and pedestrians. (In contrast, collision warning systems for personal vehicles are designed for high-speed environments and more or less constant distances to other vehicles.)

The only commercially available SODS for transit buses entered the market in 2004. Designed to operate at speeds below 15 mph and with lane changes greater than 15 mph, the SODS detects objects on the side of the bus through ultrasonic transmitters and receivers, and alerts the driver with visual and audible warnings. The U.S. DOT sponsored an independent evaluation of the SODS as deployed in transit buses in the fleets of three cities (District of Columbia, Cleveland, Ohio and Salt Lake City, Utah) to determine the system’s user acceptance and return on investment.

Lessons Learned

An independent evaluation of a Side Object Detection System (SODS) for transit buses analyzed quantitative and qualitative data to determine its Return on Investment and the extent of operator acceptance. In coordination with the participating transit agencies, the evaluators examined collision records and the costs of acquisition, training and maintenance, and assessed operator acceptance with surveys, interviews and focus groups of transit agency staff. The evaluators also conducted ride-alongs on buses to observe the system as it functioned in the field and interacted with the operator. This comprehensive approach to evaluating SODS revealed lessons learned in regards to the design and deployment of object detection systems, as follows:

  • Design a side object detection system that will detect objects in the operator's blind spot (e.g., on the far rear sides) and in locations that are frequent points of collisions (e.g., the side mirrors). The position and range of the sensors in this deployment did not allow the detection of objects near the side mirrors or at the far rear sides of the bus. This limitation significantly diminished the potential of the SODS to assist operators in detecting objects in the areas which are a frequent point of collision contact. In this regard, the system missed an opportunity to assist operators in two main problem areas. Feedback operators showed that operators considered this limitation a significant one. In fact, many operators suggested moving the sensors closer to the rear wheels and placing additional sensors on the mirrors.
  • Design visual alerts that are comprehensible and not visually distracting. Feedback from the transit operators suggests that the SODS visual alerts and other symbols were not well understood or intuitive and were distracting. Further, the visual alerts were not presented quickly enough for the operator to take action or respond. The visually distraction led many operators to disconnect, cover or turn away the display.
  • Issue audible alerts only when necessary. It is a challenge to find the balance between an alert that conveys a sense of urgency but does not annoy the operator or become a "nuisance" alert. Helping achieve this balance is to issue an alert only when it is necessary. In this deployment, the auditory alerts may have occurred too frequently to be effective. In fact, many operators revealed that they ignored the alerts over time because of the high number of false alarms. Other operators expressed concern that the alerts annoyed the passengers. Although operators said that the alerts had the potential to help them identify hazards, many expressed a strong dislike for them.
  • Deploy technologically mature collision avoidance systems. Deploying an immature system in an operational environment raises a significant risk of poor operator acceptance that can be difficult to overcome. In this deployment, the system developed technical problems and had limitations that hurt its effectiveness and diminished operator acceptance. For example, the system had a tendency to alert operators to non-hazards (e.g., snow or rain) and it did not operate consistently over time. Further, it did not detect objects in a critical area but did provide coverage in an area that operators did not particularly require. Additional research and development prior to deployment may have addressed some of these issues.
  • Engage transit operators in the deployment process. Transit operators and maintenance staff have access to knowledge that is relevant to the design and installation of collision warning systems. Their experience working in the operating environment and first-hand knowledge of the job requirements and constraints have the potential to strengthen design and deployment decisions. By engaging operators early in the process, input from the staff could inform design and deployment decisions. In addition, by involving operators earlier in the process, operators will be better informed of the goals and steps in deployment. In this case, many operators did not have a good understanding of the system’s capabilities and limitations (e.g., it cannot detect pedestrians), which led to skepticism of its capabilities and a tendency to discount it.
  • Provide training on the installation and use of the system that is consistent across sites. Training is an essential component to a successful deployment because it enables operators and mechanics to use the system effectively and understand its limitations and capabilities. In addition, the training for installation varied from site to site, resulting in uneven installations across the sites.


The evaluation of this deployment suggests that Side Object Detection Systems have much potential to reduce the number of side collisions for transit buses, and that transit agencies and operators are interested in using collision warning systems. However, the results of the evaluation point to the importance of the design and testing process to develop effective visual and auditory displays, select and place sensors that cover the areas most needed, and conduct field evaluations to ensure reliability and effectiveness in the operating environment. The lessons above are applicable to the design and deployment process and could help improve the acceptance and effectiveness of collision avoidance systems.

Side Object Detection System Evaluation: Final Evaluation Report

Side Object Detection System Evaluation: Final Evaluation Report
Publication Sort Date
12/15/2008
Author
J. Rephlo, S. Miller, R. Haas, H. Saporta, D. Stock, D. Miller, L. Feast, B. Brown
Publisher
USDOT

(Our website has many links to other organizations. While we offer these electronic linkages for your convenience in accessing transportation-related information, please be aware that when you exit our website, the privacy and accessibility policies stated on our website may not be the same as that on other websites.)

Goal Areas
System Engineering Elements

Focus Areas Taxonomy: