AI tools for prescriptive maintenance
| Reference number | |
| Coordinator | Högskolan i Gävle - Högskolan i Gävle Akademin f teknik & miljö |
| Funding from Vinnova | SEK 356 000 |
| Project duration | November 2024 - November 2025 |
| Status | Completed |
| Venture | Advanced digitalization - Enabling technologies |
| Call | Advanced and innovative digitalization 2024 - one-year projects |
Important results from the project
The objective of the project was to use simulation and digital twins to evaluate AI-generated code for machine condition monitoring and prescriptive maintenance. Within the project, an AI-generated code at the edge level is used to measure and monitor mechanical wear. The method has been verified in both simulated and real environments and is now in industrial production. Simulated results of prescriptive maintenance show that we can control production to achive full service life of the machine.
Expected long term effects
The results with prescriptive maintenance are interesting from an academic perspective, but further research is needed before it will be available in production. It is a novel research area and only few applications have been published; here the project has contributed. Generating code with AI is a highly topical topic. The experiences that the project has provided will be an asset for future projects.
Approach and implementation
The project was carried out in close collaboration between partners. In addition to project management, the project was carried out in four work packages. The project started by building a realistic simulation environment, after which methods for predictive maintenance were developed and verified. The AI generated the code, which was first verified in the simulated environment and then regenerated for the edge solution. Finally, the functionality was verified in a real environment.