PhD Forum: Trust-Aware Routing of Human Drivers to Mitigate Traffic Congestion

Route recommendation systems provide a promising way to mitigate traffic congestion in vehicular networks by recommending system-optimal paths to drivers that aim to optimize congestion at the network level. However, selfish drivers have aversions to individually sub-optimal paths and exhibit varying levels of trust towards route recommendation systems. In my previous work, we developed a novel, trust-aware Stackelberg routing algorithm that sends path recommendations to all drivers in a vehicular traffic network in an effort to mitigate overall network congestion. We then compared this algorithm to several other Stackelberg routing algorithms to demonstrate the superiority of my proposed algorithm in simulations of real-world networks of various sizes and for diverse traffic demands. In my future work, we will consider a method of modeling and learning the trust of each driver in the network from the perspective of a central routing system through the path choices taken at each routing interaction.