A Proposal to Detect Anomalies in Connections Installations using Mobile Edge Computing with Deep Learning Approaches

Deep Learning (DL) methods have been applied successfully in Smart Grid systems for the Industry 4.0 era (SGI 4.0). On the other hand, supervising the installation of connectors in the low voltage network distribution is challenging. This service requires hours of human supervision, and its maintenance generates unnecessary costs for the company. Also, there are direct impacts on the energy supplied to the consumer when a poor installation is performed. In this work, the concept of a mobile inspection system based on DL technology to detect anomalies in connector installations automatically is presented. Besides a brief literature review to support this initiative is also presented. This proposal is part of the Connection Quality initiative in the Urban Futurability project developed by ENEL Distribuição São Paulo, aiming to improve the reliability and quality of the SGI 4.0.