Predictive maintenance (PdM)is one of the most funded use-cases of Artificial Intelligence across all sectors of the manufacturing industry. The implementation of predictive maintenance techniques in industrial environments leads to efficient and sustainable manufacturing processes. In particular, it helps to the minimization of machines’ downtime (due to the appropriate and cost-effective maintenance schedule), the avoidance of unnecessary equipment replacement due to the correct detection of the root cause of a failure, and the productivity/quality increase.
Towards this direction,researchers of the Centre of Excellence for Advanced Manufacturing Technologies in IIT Kharagpurled by ProfAkhilesh Kumarand Tata Metaliks collaborated to develop a predictive maintenance algorithm able to improve the profitability of the production process and minimize downtime, costs, and manpower.
This data-driven approach has been deployed on a gearbox of an annealing furnace, and it was successfully tested for its ability to cluster various operating modes of the gearbox and outliers, detect possible failures and monitor the industrial equipment’s condition. According to the results, it is estimated that the revenues of the manufacturing process can be increased by 1 million dollars annually while the machines’ downtime could decrease by 40 hours per year as well as the requirements for manpower up to 400 hours.