Abstract: “Preventive maintenance and run to failure techniques of bearing condition monitoring could be quite expensive. The failure could equally be catastrophic thereby leading to the damage of several other components of the machinery. Many bearing studies are done at constant loading and speed and constant conditions. However, this study is carried at varying loads and speed conditions. This paper therefore presents a condition monitoring methodology based on Bayesian linear regression technique which is at varying load and speed condition.”

Keywords: condition monitoring, Bayesian linear regression, slow rotating bearings