Sharing diverse information gets driver agents to learn faster: an application in en route trip building

With the increase in the use of private transportation, developing more efficient ways to distribute routes in a traffic network has become more and more important. Several attempts to address this issue have already been proposed, either by using a central authority to assign routes to the vehicles, or by means of a learning process where drivers select their best routes based on their previous experiences. The present work addresses a way to connect reinforcement learning to new technologies such as car-to-infrastructure communication in order to augment the drivers knowledge in an attempt to accelerate the learning process. Our method was compared to both a classical, iterative approach, as well as to standard reinforcement learning without communication. Results show that our method outperforms both of them. Further, we have performed robustness tests, by allowing messages to be lost, and by reducing the storage capacity of the communication devices. We were able to show that our method is not only tolerant to information loss, but also points out to improved performance when not all agents get the same information. Hence, we stress the fact that, before deploying communication in urban scenarios, it is necessary to take into consideration that the quality and diversity of information shared are key aspects.