Automated traffic enforcement which respects "driver privacy"

At many intersections in downtown Los Angeles, cameras take pictures of the license plates of vehicles which run red lights. This information is used to automatically mail tickets to offenders. It seems likely that such systems becomes increasingly widespread in the near future, as they are perceived to be effective in cutting down on traffic infractions, boosting governmental revenues, and freeing law enforcement officials for other tasks. However, a network of cameras poses dramatic threats to the privacy of motorists - if kept continuously running, such a network could be used to track detailed driving habits of specific individuals. Speculative proposals involving state-monitored GPS transmitters in every registered vehicles pose even greater threats to "driver privacy". This situation is an example of the standard tension between citizen privacy and law enforcement. In this paper, we discuss the design of a novel camera-free protocol for traffic monitoring involving EZ-pass-like transponders. Our system uses cryptographic algorithms to guarantee that the state is able to detect and identify violators of a wide variety of traffic laws (light violations, speeding, illegal turns, and so forth) without having the ability to track the movements of motorists. That is, we design a traffic monitoring system with the property that no matter how it is misused, it is impossible to reconstruct the driving paths of specific cars. On the other hand, the system also guarantees that specific information about motorists who commit traffic infractions can be recovered. In addition, this protocol can be easily adapted for sophisticated and nuanced anonymous toll-collection. The system is fairly easy to implement and is immune to a variety of attacks and cheats.