Safety issues in collaborative vehicle control

Abstract

Collaborative driving systems are complex, distributed systems that control physical vehicles. They can reduce accidents, decrease fuel consumption, reduce commute times and increase the capacity of highways. However, errors in their implementation can cause unsafe conditions that result in the loss of human life. Our objective is to find ways to design, verify, and quantify safety in these systems. In this paper we describe a collaborative driving system that assists in safely and quickly merging vehicles when highways merge, following tolls, and at construction or accident sites. We describe 1) an architecture that partitions the application into modules that can be tested and modified independently, 2) a communication protocol that provides an unconventional set of services that simplifies the implementation of the system, and 3) a strategy for cooperation between the human operator and the system.