MACHINE LEARNING FOR DECISION-MAKING IN THE REMANUFACTURING OF WORN-OUT GEARS AND BEARINGS

Mechanical industries use rotating mechanical equipment in their day to operations. The suffers from wear and tear, is usually discarded as scrap. But there a way recover some of this reuse it? This paper uses machine learning capture analyse the wearing damage bearings gears determine whether they can be redeemed. Finite element analysis conducted on worn-out spur pillow order facilitate feature extraction image processing algorithms. converts actual gears, bearings, seals into CAD files. decision-making system designed, it these files decide optimum manufacturing process restore redeemable components. components are designed using SOLIDWORKS. MATLAB, Proteus software, Arduino micro-controller used for application design simulation. results tests gear bearing show that 4% non-redeemable, while 60.2% non-redeemable. decision taken by redeem discard bearing.