Coordinating mobile robot routes with machine schedules is a complex shop floor challenge. A multi-agent reinforcement learning framework links these decisions and demonstrates how robots can adapt to processing delays and equipment failures in simulation.