Development of a Holistic Framework for Identifying Energy Efficiency Potentials of Production Machines

A prerequisite to identify energy efficiency potentials and to improve energy efficiency is the measurement and analysis of the energy demand. However, in industrial practice, approaches to identify energy efficiency measures of production machines are associated with high costs for metering equipment and time consuming analysis requiring expertise. Against this background, this paper describes a comprehensive and cost-efficient framework from acquisition to analysis of energy data to serve as a starting point to increase energy efficiency in manufacturing. For this purpose, an energy transparency and analysis system is being developed that can measure, record and analyze electrical quantities. The validity of the data acquisition can be verified by utilizing a Raspberry Pi as a low-cost edge analyzer device. Measurement data is stored with associated metadata in a SQLite database for subsequent processing in a Python-based web application, in which machine learning algorithms can be deployed. The algorithms can be used to process vast amounts of data and to provide a basis for calculating energy performance indicators to reveal energy efficiency potentials. The overall workflow is validated using a lathe and a cleaning machine within the ETA Research Factory at the Technical University of Darmstadt