Sustainable Fog-Assisted Intelligent Monitoring Framework for Consumer Electronics in Industry 5.0 Applications
Subhranshu Sekhar Tripathy, Sujit Bebortta, Thippa Reddy Gadekallu
Abstract
The fifth era of the industry (Industry 5.0) has been marked by the reformation witnessed in consumer electronics sector by bringing forth technology that could enhance efficiency, connectivity, and user experience. Industry 5.0 makes it possible to create intelligent consumer electronics products that can interact, analyse data, and instantly adjust to user preferences. Fog processing further enhances Industry 5.0 by bringing processing power closer to end-user devices at the network’s edge. Traditional machine learning techniques are unsuitable for manufacturing use cases which demand high degree of interoperability and heterogeneity due to the unavailability of private data, which requires decentralized learning solutions. To address this, we designed a monitoring framework that uses deep reinforcement learning to predict the effect of mobile computing resources in manufacturing systems and detect disruptions in real time. Our framework is deployed at the Fog computing level and includes a dynamic rescheduling module that sustainably optimizes task assignment, improves execution accuracy, reduces delay, and maximizes the resource utilization. Numerical results demonstrate the efficiency of our scheme in managing task rescheduling and real-time disruption detection, depicting the sustainable utilization of available resources over the considered benchmark algorithms.