Development of a Peer-to-Peer Collision Warning System Qingfeng Huang , Ronald Miller , Perry MacNeille , Gruia-Catalin Roman , David DiMeo Department of Computer Science, Washington University, Saint Louis, MO 63130 Distributed Intelligence in e-Technology, Ford Research Lab, Dearborn, MI 48124 qingfeng, roman @cs.wustl.edu rmille47, pmacneil, ddimeo2 @ford.com ABSTRACT Traditional vehicle collision warning and avoidance systems do not perform well in the perpendicular path intersection case. One major reason is that the threat detection systems they use require line-of-sight. Recently we have designed and implemented a new low-cost collision warning system that doesn’t require line-of-sight for threat detection. The novel elements of the system include the use of a dynamic ad hoc wireless network for peer-to-peer data sharing, a new intersection collision warning algorithm, and a flexible and extensible software architecture and system design. This pa- per, the first paper report on this work, focusing on the sys- tem design elements. Detail of field test results and perfor- mance tuning will be reported at a later date. Keywords Vehicle Collision Warning System, Ad Hoc Networks, Spec- ification, Algorithm 1 Introduction There have been major improvements in vehicle safety since the 1960’s. The introduction of safety features such as seat belts, air bags, crash zone, lighting and new vehicle struc- tures has dramatically reduced the rate of crashes, injuries and fatalities. The fatality rate per hundred million miles travelled has fallen from 5.5 to 1.7 in the period from mid- 1960s to 1994 [7]. However, in spite of these impressive improvements, each year in the United States, motor vehi- cle crashes still account for a staggering 40,000 deaths, more than three million injuries, and over $130 billion in financial losses. All of these safety features are either static or passive. They act to minimize collision damage or give the driver vi- sual assistance or warning at specific geographic areas. With recent advance in sensing, computing, and communica- tion technologies, researchers have started to design and de- velop more advanced systems to further improve automobile safety. New driving assistance systems such as night vision systems and collision warning systems (CWS) have been de- signed, tested, and deployed [7, 3, 4, 2, 1]. While night vi- sion systems simply provide visual assistance to drivers in dark environment, collision warning and avoidance systems generally exhibit some intelligence. By actively monitor- ing vehicle surroundings and the driver’s state, these systems warn the driver of hazard, allowing drivers to take appropri- ate actions to avoid an accident or to reduce the severity of the crash. Preliminary results have shown that the introduc- tion of collision warning systems could dramatically reduce crash fatalities, injuries and property damage [7]. Studies carried out by Daimler-Benz and National Highway Traffic Safety Administration (NHTSA) suggest that additional one second warning could reduce the rear-end and intersection accident rate by 50 to 90% [5], and Eaton reported that the actual truck fleet accident frequency was reduced by 73% on fleets being equipped with VORAD Forward and Side Colli- sion warning systems by Eaton [5]. Despite the fact that intersection collision accounts for al- most 30% of all crashes, intersection collision avoidance systems received less attention than the forward collision avoidance systems [3, 6]. The reason, besides the fact that the intersection collision problem is more complicated than rear-end crash, is the limitation of the radar technology, the most widely used object sensing method in vehicle collision avoidance systems. Most radar systems require line-of-sight for object detection. Yet in most intersection crash cases, the principle other vehicle (POV) is hidden from the line of sight of the subject vehicle (SV) until the last second before the collision. This renders ineffective most collision warn- ing/avoidance system that requires line-of-sight for threat de- tection. Recently we have designed and developed a system capable of intersection collision warning using a new approach. The system is based on vehicle-to-vehicle communication on top of ad hoc networks. Threat detections are achieved by vehi- cles cooperatively sharing critical information for collision anticipation, i.e., location, velocity, acceleration, etc. By sharing the information between peers, each vehicle is able to predict potential hazard. Although this system doesn’t re- quire a support infrastructure, the ultimate value of this kind of peer-to-peer cooperative system depends on the percent- age of vehicles on the road using it. The more vehicles use it, the more valuable it is. Research is underway to estimate the critical mass needed for the system to have a substan- tial value gain. As more and more vehicles are equipped with navigation and communication systems, we envision a near future when this type of system will demonstrate its ad- vantages in public use, complementing the function of other