This thesis seeks to investigate a long-running system that handles truck platooning in a fuel-efficient manner while also able to handle new assignments. A long-running system is a system running continuously, receiving and executing request after request, while keeping its internal state up-to-date. As trucks start their journey, they send a request to the coordination system, registering with the system. The coordination system keeps track of the position of all the trucks, and can update the plans in the system.

Methodology

A simulation engine was created to investigate the viability of a long-running system, and the performance of different algorithms used in the system. For evaluation, several large test sets of truck assignments were randomly generated to model real world transport assignments. These truck assignments were then used in the simulation to investigate the effect of systematic parameter variations on the obtained fuel consumption reduction.