Toward Digital twin for sustainable manufacturing: A data-driven approach for energy consumption behavior model generation

Traditional manufacturing systems face a major transition challenge toward intelligent and sustainable manufacturing systems. Energy-efficient manufacturing has become one of the main focuses of this transition because of the high and critical consumption of energy for the production of many industrial sectors. To this aim, modeling Energy Consumption (EC) behavior is a primary step to monitor and reduce consumption. However, it requires a deep understanding of the system's kinematic and dynamic behaviors. Hence, creating a model from scratch can be challenging, which may result in a model that does not correctly represent the real system. With the advancement of digital technologies, it is now possible to collect and analyze data from manufacturing systems in real-time. This opens the door to the possibility of modeling the energy consumption behavior of different machine states based on a data-driven approach and keeping the current energy consumption under control and monitored using real-time data from the equipment. Prior research has been conducted in the literature incorporating EC modeling into a Digital Twin (DT). However, the addressed issue remains an open challenge due to its complexity. The proposed methodology solves the lack of literature by proposing a meth-odology that makes the EC modeling within the reach of any researcher or practitioner in the field. The current paper proposes a data-driven methodology for integrating the EC model into Digital Twins. The methodology is based on measurements to identify different segments and sub-states of EC of production equipment, using techniques such as segmentation and regression. It relies on power absorption measurement of industrial equipment to generate EC related-parameters to be fed into the DT model to monitor the current operating condition of the physical system. This work contributes to the DT-based sustainable transition by allowing to monitor and quantitatively measure those parameters which could be controlled to reduce EC. A case study on an industrial robot is used to validate and assess the performance of the approach in a laboratory environment.