The paper discusses the design and implementation of a cognitive digital twin (CDT) to enhance the capabilities of autonomous industrial mobile manipulator (AIMM) in industrial settings. The integration of data in CDT facilitates enhanced positioning, even in the presence of obstacles, and optimized recharging schedules. The use of external sensors significantly improves the robot’s accuracy. The implementation of ML models facilitates intelligent planning of charging stops and minimizing downtime. This enhancement is achieved by creating a virtual representation of the physical system that incorporates cognitive capabilities, such as reasoning, learning, and planning. The study demonstrates the potential of CDTs to serve as advanced tools for decision-making, optimization, and predictive maintenance in industrial settings. The results also highlight the challenges in developing CDTs, particularly the need for high fidelity in replicating the physical system and the environment.