Session Outline
In this project, SE a.s. is implementing an AI process to support defect classification performed by screening committee and save manual labour time. The scoring engine is based on a series of neural networks and natural language processing methods with binary and multi-label outputs. We predict severity, cause type and impact area. Predictive model inference is performed in real time at the moment of filling into the maintenance system and predicted classes are assigned according to specified probability thresholds.
Key Takeaways
- Application of AI methods in maintenance processes
- Example of Natural Language Processing usage in energy industry
- Improve safety in nuclear industry
Add comment