Description
We describe a decentralized routing algorithm with physical time windows for modular conveying systems. Existing routing algorithms for modular conveyors are already capable of bi-directional conveying while avoiding conflicts such as collisions, deadlocks, livelocks and starvation effects. In addition to avoiding conflicts, routing algorithms must also select routes that minimize the transport time. No existing algorithm for modular conveyors bases this decision on the expected physical lead time, even though physical lead time directly affects the system throughput. In this publication, we present an algorithm that uses the physical lead time to select routes while avoiding conflicts. The avoidance of conflicts is mathematically proven and the algorithm's computational complexity is calculated. We present the system behavior of an exemplary layout which consists of nine modular conveying modules that are controlled by our algorithm. With only nine modules, the package throughput is on the same level as the package throughput of conventional sorting systems. Due to its modular design, additional modules can be added to further increase the throughput, thus surpassing the throughput of conventional sorting systems.