Scheduling mobile robots in part feeding systems

Purpose: Industry 4.0 has increased the availability of real-time data in manufacturing systems, but scientific evidence about the value stemming from such data is still lacking in several fields. This paper studies data-driven approaches for the assignment of tasks to a fleet of mobile robots transporting parts to the stations of a mixed model assembly line. The approaches exploit real-time data concerning the robots and assembly stations state. Methodology: An agent-based simulation model of the system, including factory warehouses, assembly stations, and robots, is developed and validated through a real case in the automotive industry. Findings: The paper proposes a model that measures the part feeding system performance in terms of transportation tasks completion time, idle time of the assembly stations due to lack of materials, and amount of inventories at the assembly line. Different data-driven approaches are considered, differing among each other for the type of real-time data used and for the update frequency of the task assignment. Originality: The developed model enriches the ones presented in previous literature by including new information (e.g., robots failures) and new data-driven approaches, such as the dynamic assignment of tasks to robots.