Metaheuristic algorithms for improved optimization of the distribution of end-of-life dismantling tasks between humans and robots

The wide use of robots in disassembly is an important milestone toward moving to more efficient and sustainable product lifecycle management. The focus is on conserving resources and reducing waste. Despite the progress made in this field, the presence of a human operator is still required to perform tasks outside the outreach zone of the robots. Therefore, a perfect synchronization and complementarily between humans and robots is crucial. This work deals with the planning of disassembly tasks between the robot and the human subject to energy losses. The main objective has been to improve the efficiency of disassembly processes through intelligent techniques for optimizing both robotic and human disassembly task allocation. To this end, several optimization algorithms were implemented, mainly, Salp Swarm Algorithm (SSA), Particle Swarm Optimization (PSO), and the Grey Wolf Optimizer (GWO). A comparative analysis of the performance of these algorithms has been carried out using two industrial case studies. The performance of each algorithm subjects to evaluation criteria such as execution time, investment cost, energy consumed by the actuators, and total disassembly time has been assessed. The outcomes confirmed that SSA algorithm outperforms the PSO and GWO algorithms. The suggested tasks allocation by the SSA algorithm saves up to 25.42 % on investment costs, 73.85 % on execution time and 8.5 % on the energy consumed by the actuators which proves that the SSA solution has a better carbon footprint than the other solutions. A total of 20 runs has been performed for each technique to investigate their stability in solving such optimization problems. The outcome confirms that SSA is the most stable in terms of the quality of the results.