A Digital Twin for Condition Monitoring of Conveyor Belt Systems

Predge AB and LKAB are jointly developing an innovative solution for condition monitoring and health prediction of conveyor belt systems, called CoCoP. A digital twin is developed and validated to provide predictive decision support for operation and maintenance of conveyor belt system. The degradation of all critical components due to wear and tear including their interactions are represented in the digital counterpart. Degradation towards failure is quantified and the time point of failure in the future is foreseen by the digital twin.

Conveyor belt systems are composed of numerous components where the failure of a component leads to production stops of varying severity. While idlers may fail more often, they are usually more swiftly replaced than drums and pulleys which render longer production stops for their repair. Bearings in all these components are very often the failing part of these components. The CoCoP solution is using a digital model, combining component information with the dynamic behavior of the degradation processes, called a digital twin. The digital twin will operate in parallel with the real-life conveyor belt and continuously monitor and assess each component based on operational and maintenance data. In addition, visual information from the operation is used by CoCoP to determine extraordinary hazards to the operation of conveyor belt using an AI approach.

For the condition monitoring of the components of conveyor belt system models with predictive capabilities are derived using a combination of first principle models and data-driven models (methods from both machine learning and system identification are used here). Idlers, rollers, and pulleys are then monitored, and the condition is predicted to establish estimates for the remaining useful life of individual components. Maintenance can then be appropriately planned and optimized by assessing the future condition of the population of components. A decision support will enable the different roles like operators, maintenance engineers, and management to act on the provided insights. The decision support also includes a scenario-based assessment tool which uses the digital twin to perform what-if analysis

The project consortium consists of LKAB and Predge AB, where LKAB provides access and expertise on the facility for validating solution, and Predge AB provide the digital platform and the prototype. The joint effort targets a market-ready solution that increases utilization and reduces unplanned stops. An external reference group with experts on conveyor belt systems from Boliden and different sites of LKAB are following the project’s progress and provide recommendations and advise to the development team. This will enable efficient replication of the CoCoP solution to other conveyor belt system within LKAB or Boliden, but also to third parties.

Both LKAB and Predge have been collaborating in several research and development project and have ongoing commercial contracts. The project is partially funded by EIT RawMaterials.