Path planning algorithm for levitating planar motion system

Sammanfattning

In today’s world of ever increasing competitiveness, solutions that include automation and smart production have become a vital part to consider in overall business strategy, specifically for the industry sector. Whether an automated production process will be beneficial or not is dictated by how sub-processes such as interproduction transportation operates. The importance of these processes has been displayed in the late increase of production efficiency when moving from traditional transportation units, such as conveyor belts, to more sophisticated systems, such as transportation robots. However, these new sophisticated systems comes with increased complexity and new challenges when implementing important behaviours such as speed, control and safety. This thesis is linked to the challenge of developing a safe and time efficient feature for handling a sudden failure or halt in one of these systems, namely the Beckhoff XPlanar levitating planar motion system. Hence, the goal of the thesis was to develop a pathfinding algorithm to easily line up the agents in the XPlanar system from any given position to a pre-specified startup track. The end-result was a multi-agent pathfinding algorithm that utilizes Conflict-Based Search and A* to move each agent from their start position to a desired end-position whilst avoiding collisions. The algorithm is specifically designed towards the XPlanar system, integrated through ADS communication making it executable from the Beckhoff PC-based control software TwinCAT3.