The goal of Predictive Maintenance (PdM) is to identify approaching equipment failures and conduct Preventive Maintenance (PM) before the failures occur.
PdM is advantageous over periodic scheduled PM because PdM allows to conduct PM only when necessary, thereby reducing PM cost and asset downtime.
Many PdM companies emerged in recent years due to reduced cost of sensors and IT, as well as improvements in Machine Learning capabilities.
Asset owners are faced with many options for PdM systems, and selecting the right one is a challenge.
While PdM offers many advantages to asset Operation & Maintenance (O&M), PdM systems are not cheap.
The cost of PdM includes: installation of sensors and IT, data storage and servers, software updates, and usually a data scientist.
In many cases assets already have a lot of collected data (from SCADA and EAM) that is not utilized, and new expensive PdM systems are not necessary.