Real-time coordination of autonomous vehicles

Autonomous vehicles seem to be a promising approach to both reducing traffic congestion and improving road safety. However, for such vehicles to coexist safely, they need to coordinate their behaviour to ensure that they do not collide with each other. This coordination is typically based on (wireless) communication between vehicles and needs to satisfy stringent real-time constraints. However, realtime message delivery cannot be guaranteed in dynamic wireless networks which means that existing coordination models that rely on continuous connectivity cannot be employed. In this paper, we present a novel coordination model for autonomous vehicles that does not require continuous real-time connectivity between participants in order to ensure that system safety constraints are not violated. This coordination model builds on a real-time communication model for wireless networks that provides feedback to entities about the state of communication. The coordination model uses this feedback to ensure that vehicles always satisfy safety constraints, by adapting their behaviour when communication is degraded. We show that this model can be used to coordinate vehicles crossing an unsignalised junction