While the emerging self-driving vehicle technology is expected to improve both the safety and efficiency of transportation, such improvements are not guaranteed or even possible if behaviors of automated vehicles (AV) are not coordinated with other vehicles (automated or manually driven) on the road. The challenge of coordination is even greater when a large-scale network of vehicles and a mixed environment of manned and autonomous vehicles are considered.

Utilizing emerging vehicular communication technologies, it is possible to develop methods for the coordination of actions of a large-scale network of connected and automated vehicle (CAV) systems. This research aims to push the boundaries of coordinated operation of a large network of CAVs in a mixed environment of manned and automated vehicles, through the concept of perceptive stochastic coordination. This concept targets the mass platoon control problem by modeling vehicle movements and utilizing the models in a Stochastic Model Predictive Control (SMPC) setting. We explored the performance of the vehicle platoon with different information flow topologies, where vehicles receive information from multiple predecessors, in both ideal and non-ideal communication setups.