Multiagent Systems in Automotive Applications

Abstract

The multiagent systems have proved to be a useful tool in the design of solutions to problems of distributed nature. In a distributed system, it is possible that the data, the control actions or even both, be distributed. The concept of agent is a suitable notion for capturing situations where the global knowledge about the status of a system is complex or even impossible to acquire in a single entity. In automotive applications, there exist a great number of scenarios of distributed nature, such as the traffic coordination, routes load balancing problems, traffic negotiation among the infrastructure and cars, to mention a few. Even more, the autonomous driving features of the new generation of cars will require the new methods of car to car communication, car to infrastructure negotiation, and even infrastructure to infrastructure communication. This chapter proposes the application of multiagent system techniques to some problems in the automotive field.

An Agent-based Approach to Continuously Detect and Update Road Network Changes Using GPS Trajectories

“…The main objective of agent-based systems is to develop autonomous systems capable of cooperating with systems in the real world [34]. The agent-based systems are composed of multiple autonomous agents that work together to achieve an objective that could not be accomplished by a single agent acting alone.…”