Developing predictive maintenance algorithms involves a number of challenges. Using the example of a packaging machine, the article explains how an algorithm for predictive maintenance can be developed.
Predictive maintenance has become essential in industry to increase operational efficiency and reduce maintenance costs. However, developing predictive maintenance algorithms for vehicles, MRI equipment, wind turbines or assembly lines also poses some challenges to the engineers involved. Algorithm development requires not only extensive experience with machine learning techniques, but also a solid understanding of asset behavior. However, professionals who exhibit skills in both areas are hard to find.
Implementing functionality on the shop floor also requires a series of steps to deploy the predictive maintenance algorithm on an industrial controller, edge device or in the cloud. Parts of a single algorithm can run on different parts of the infrastructure, further increasing complexity.
Using a packaging machine as an example, this article shows how a predictive maintenance algorithm can be developed and deployed in a production system.
The exemplary packaging machine is equipped with several robot arms. These move back and forth at high speed when placing objects to be packaged on the assembly line. They are connected to industrial controllers (PLCs) that communicate with a Microsoft Azure-based system. This IT/OT system collects streaming data from the edge devices connected to the robotic arms and runs predictive maintenance algorithms based on that data. This enables the detection of anomalies and prediction of potential failures of the arms, and reports the results back to dashboards used by engineers and operators.