Optimizing energy consumption of robotic arm movements based on Digital Twins

Robotic arms are extensively used in production environments to undertake tasks such as welding, hemming, etc. Minimizing energy consumption of robotic systems poses a critical challenge for sustainable manufacturing. We propose using the robot’s Digital Twin to obtain the energy consumption for each movement of a production cycle under different operational scenarios, i.e., different configurations for attributes such as velocity, acceleration, jerk and trajectory. Further, we develop an Integer Programming (IP) model that incorporates these scenarios and selects the ones that minimize total energy consumption. To facilitate applicability, we present a preprocessing filter that uses Pareto dominance to remove suboptimal scenarios, reducing the IP’s solution space and thus vastly improving computational efficiency, as also shown in our computational experiments. Moreover, we present how we seamlessly apply our approach within the design process of robotic cells.