Digital Twin-Based Imitation Learning for Human-Like and Efficient Driving of Mobile Robots in Virtual Factory Environments

Abstract In smart manufacturing and flexible production systems, autonomous mobile robots must complete logistics tasks efficiently while exhibiting driving behaviors that human operators can interpret and accept. This study proposes a simulation-based digital-twin imitation learning framework for learning human-like driving patterns in virtual factory environments. The proposed system is built on the proximal policy optimization algorithm and integrates generative adversarial imitation learning to establish a dual-reward structure that combines extrinsic rewards with demonstration-derived intrinsic rewards. Evaluation was conducted on seven virtual factory routes, including loop, test, serpentine aisle, junction, narrow U-turn, loop-short cut, and asymmetric layouts. The study also compares the method with behavior cloning and DAgger and analyzes intrinsic reward strength at five different levels (0, 0.001, 0.1, 0.5, and 1.0). To rigorously evaluate performance, operational metrics, such as completion rate, cycle time, collision count, shortcut use, path length, and an energy proxy, were measured, while the dynamic time warping algorithm was used to quantify the similarity between human trajectories and the AI agent’s paths. Experimental results demonstrate that stronger imitation rewards can improve trajectory similarity on several maps, but they also introduce efficiency and robustness trade-offs across route topologies. This research provides a methodology for pre-training Physical AI in a simulation-based environment, offering a tunable approach to balance task efficiency and human-like behavior before actual field deployment.