Cyber Physical Regional Freight Transportation System

The sharp increase of e-commerce over the last few years has led to an augmentation in the volume of trucks both in ports and in commercial areas. Truck traffic has a negative impact on traffic flow in general, due to the size of trucks and their slower dynamics. The continuously increasing use of navigation apps has led drivers to make their routing decisions in an independent manner in an effort to minimize their own individual travel time, with possible significant deviation from a socially optimum solution. On the other hand, in a socially optimum solution, some drivers may benefit but some others may get harmed compared to the situation where they make their decisions in a selfish manner. To alleviate these issues, we designed a mechanism whose main purpose is the creation of individual incentives for voluntary participation of the truck drivers with simultaneous minimization of a socially optimum cost. More specifically, users send their OD matrices as well as their preferred departure time to the coordinator who gives them routing instructions based on a socially optimum criterion. This design enables us to derive sufficient conditions under which we prove the existence of truthful in equilibrium and budget balanced on average mechanisms, which create individual incentives for voluntary participation of the truck drivers. Subsequently, we designed our mechanism in a way that it only uses a minimal set of sufficient conditions in order to guarantee the existence of a solution and maximize its efficiency. The extensive simulation results of our approach in the Braess and the Sioux Falls networks demonstrate that the proposed mechanism can approach the system optimum solution.

The optimum routes are found by using a co-simulation approach where a real time traffic flow simulator generates travel times and travel costs which are fed into an optimization block to generate optimum routes by  minimizing an overall cost. The generated routes are then tested using the simulator and the load balancing block makes adjustments to the distribution of freight along different routes in order  to balance the effect of the loads on travel times and link flows across different routes. Since effect of loads on travel time and link flows is a nonlinear unknown function the process becomes iterative till certain criteria are satisfied.