Motion planning optimization using hierarchical algorithms for multi-robot cells

In manufacturing, motion planing of multi-robot cells is a demanding activity. The issue of work-piece positioning is still mostly solved using the human competence, by trial and error procedures, making feasible solutions hard to find, and making it a time consuming procedure. Surely, the optimization of a custom objective function along a path is complex, due to the dimension of the feasible-configuration space, its high non-linearity, and to the collision probability along complex trajectories. Furthermore, a human approach to the problem, makes the exploit of the task redundancy, often found in industry, a complex matter. This work proposes a twolayer iterative optimization method that integrates an external Whale Optimization and an internal Ant Colony Optimization algorithm, allowing the optimization of a user-defined objective function along a working redundant path, to produce a quasioptimal, collision free solution within the feasible-configuration space. The trials result a good trade-off between computing time and solution optimality.