LAUSR

Abstract The article presents a risk level assessment for rear-end collision which depends on interactions between environment (e.g. cars, pedestrians, road etc.), ego driver and vehicle. The evaluated risk focuses on the probability of collision between ego and front vehicle. Information comes from embedded sensors which provide data about inter distance, vehicle dynamic, temperature etc. Our system has also communication capabilities, i.e. vehicle to vehicle (V2V) communication, which allow the ego car to sense his environment. Then it infers a probability of risk with a Bayesian Network. In a previous work this risk level assessment tool was developed for the worst case assumption concerning the inter distance. By adding an estimation of the braking intention of front car, the tool presented in this paper work on a less restrictive assumption. As presented this tool is an ADAS for human conducted vehicles but it can be easily adapted for autonomous car.