Artificial Intelligence for Optimal Truck Platooning: Impact on Autonomous Freight Delivery

Abstract

The advancements in autonomous- and connected-vehicle technologies bring drastic changes in freight delivery. Vehicle-to-vehicle and vehicle-to-infrastructure communication has become a reality with the help of autonomous and connected vehicles. One of the most notable changes is the formation of truck platoons. Despite the numerous benefits of truck platooning, such as reduced fuel consumption and increased traffic efficiency, this approach requires a significant amount of computational resources to obtain aerodynamic performance under different scenarios. To overcome this challenge, a data-driven surrogate model was proposed to predict the drag force and fuel-consumption rate of truck platoons. The surrogate model improves computational efficiency, as compared to traditional methods, and provides a valuable tool for evaluating the performance of truck platoons. To demonstrate the benefits of truck platooning, a 161-km (100-mi) corridor in Illinois on I-57 highway was selected to conduct fuel-consumption analysis and delivery-cost analysis for a three-truck platoon. The results showed that the average fuel savings achieved can be up to 10%, depending on the headway between the trucks. The delivery cost of the truck platoon was reduced by 30%, as compared with conventional line-haul delivery. These findings highlighted the importance of truck platooning as a solution for reducing fuel consumption and improving delivery economy in the freight industry.