A Study on Industrial Process Optimisation and Predictive Maintenance using AI-mRL

Industrial 4.0 automation environment integrates a combination of mechanical devices, process machinery, Inductive and resistive loads that are monitored and operated through sensors, PLCs, PAC and Industrial IOT (IIOT) devices for better process and efficient operations. Creating a seamless operating environment requires a robust and most advanced technology integration and data mining environment. Our Research uses multiple fusion sensing hardware and embedded programming with Deep learning and AI algorithms to predict industrial process/machine failures using,  An Embedded AI edge device (hardware).  Long distance Sub-GHz communication protocol.  AI agent to collect, analyse and tigger faults on local and cloud infrastructure. The hardware uses a micro controller with Wi-Fi communication to collect and transmit data. The system captures data on multiple sensing elements including Power, Vibrational and temperature for smart analytics to sense and report failure mode trigger to avoid process failures and machine breakdown. The research will understand the industrial process efficiency and performance by collecting data from multiple sensors for a data driven approach using Machine learning and deep learning algorithm. “A fault trigger before the machine breakdown”, that minimise the cost and improve the process efficiency using Machine Reinforcement Learning (mRL).