A Cyber-physical Systems Approach to the Design of Vehicle Safety Networks

The concept of cooperative vehicle safety (CVS) is based on cooperation of vehicles in achieving real-time situation awareness for the purpose of safer (and possibly autonomous) driving. Since CVS is a safety system, situation awareness has to be real-time with high accuracy. Cooperation in the context of CVS is achieved through vehicular networks that are tuned to the purpose of vehicle state tracking in a neighborhood around each vehicle. In this paper, we describe different methodologies in approaching the design of CVS, and propose a systematic Cyber-Physical Systems approach to the design of CVS. Our proposed approach is to design the system in a top-down manner, by identifying system metrics such as approximate measures of awareness, and then characterizing component models and their interactions. For this purpose, processes that handle vehicle state estimation and dissemination of state information are studied and their interactions are modeled in order to derive an overall awareness quality metric. The result of analyzing and modeling such effects is then used in studying the design of adaptive control schemes for each component of CVS.