MachineHealth: Towards healthy machines and predictive maintenance with AI



Örebro universitet


Alfred Nobel Science Park, Zinkgruvan, Epiroc

Amy Loutfi



Purpose and goal
This project will conduct a feasibility study to understand the requirements for monitoring machine health in the mining environment. With the latest announcements of new technology, AI algorithms will be used. The two techniques used are deep learning algorithms and techniques that allow numerical representation to be abstracted in a way that can be easily and intuitively understood by the human operator. We will develop a concept for expanding the monitoring of machines and increasing the analysis so that the most important patterns are presented directly to machine owners and operators.

Structure and implementation
In the project we will use established AI algorithms from Örebro University together with the data available with Certiq. The preliminary study will evaluate the feasibility of tested AI algorithms to be applied to a new domain to investigate the ability to provide automated analysis and reveal trends that can not always be easily found by human operators.

Expected results
In particular, we follow the vision in STRIM, by developing completely independent mining operations and fewer hours worked. MachineHealth will primarily focus on developing the position of the technology in relation to Section Efficient unit operation for mining, by developing unit operations to facilitate automation and efficient recovery.