Trajectory Planning and Runtime Control of Planar Movers with a 2D Collision Avoidance Algorithm

Motion control for multi-agent systems is crucial for managing complex scenarios like traffic management, automated driving, and robotics. Collision-free navigation in crowded environments is essential, and systems using sensors, cameras, and AI play a key role in preventing collisions and reducing risks through warnings or active intervention. This thesis addresses the challenge of safe and efficient navigation in multi-agent systems, focusing on collision avoidance in industrial planar transport systems. The Beckhoff XPlanar transport system, using magnetic levitation for mover translation and rotation, served as the test platform for validating the proposed algorithm. Such solution implies the definition of a safety areola around each mover, enabling the computation of a related proportional damping action that slows down movers to avoid collisions without path deviations where possible. The results show the algorithm’s effectiveness in handling complex scenarios such as priority assignment and allows the use of overtaking actions when a solution cannot be found otherwise, ensuring smooth operation in multi-mover environments. This work presents the developed collision avoidance algorithms and simulation-based experimental results, demonstrating their effectiveness for decision-making, together with related drawbacks and its possible implementation in other systems.