LAUSR

ABSTRACT To comply with the trend of customised production, this paper proposes a novel methodology for a changeable logistics system on certain factory scenes. By modularising the traditional conveyor belt system and applying decentralised control to each module, the changeable conveyor system can quickly reassemble itself for different logistics routes and thus react to the changing requirements of new production plans. The logistics route transformation is organised in two stages while reassembling. First, the modules calculate their paths on their own towards the assigned positions using a modified Rapid-exploring Random Tree (RRT) algorithm with the introduction of orientation factor (α-factor) to increase path smoothness; then a sensor-based Artificial Potential Field (APF) algorithm is applied for real-time collision avoidance with the sensor orientation and data weight being specially arranged. Through these two stages, the proposed system can flexibly react to changing logistics demands and form various logistics routes in the factory layout according to customised production requirements. Simulations and comparisons are given to illustrate the performances of the method.