AI-Based Fault Detection system for Mine Dewatering Systems

The removal and movement of water is an essential activity for the safe and productive operation of mines. In the mining industry pumps are often required to operate several km underground. If a pump were to fail in such a location, the results can be costly and potentially unsafe. It can be exceedingly difficult to apply conventional on-line condition monitoring systems to warn personnel of an imminent pump failure. Technosub, the industrial partner in this work, has recognized that there is presently no off-the-shelf, commercially available solution for remotely monitoring the majority of the pumps that they sell for operation in extreme locations. To address this challenge, Technosub is endeavoring to develop a small, self-contained, hardware-based fault detection device that is mounted directly to individual pumps. If a pump is about to fail, the system will then communicate this vital information to maintenance personnel. The proposed project involves the development of the computational framework for the intelligent system that will reside on the monitoring device to determine the pump’s condition. The intern on this project will gain valuable experience developing and applying advanced technologies to an industrial application.

Faculty Supervisor:

Markus Timusk

Student:

Danyk Levesque

Partner:

Technosub

Discipline:

Engineering

Sector:

University:

Laurentian University

Program: