impact
- Leveraging the existing control system.
- Reduction of the number of sensors required to obtain the improvement.
- Use of energy consumption to evaluate mechanical behavior
Developed AI Technologies
- Online Machine Learning-based non-intrusive load monitoring.
- Multi-consumer baseline generation
Challenges
- Transformation of the current maintenance model based on catalog-based preventive (without considering actual use) and corrective maintenance (with the consequent system downtime, reduced availability and higher repair costs).
- Use of a single measurement system to obtain data from a line (30 conveyors), applying Machine Learning algorithms for start-up differences identification and degradation tracking by high-speed energy consumption measurement..
results
- Reduction of unscheduled shutdowns (>80%).
- Reduction of preventive maintenance cost (50%)
- Reduction of corrective maintenance cost (20%).
- Reduction of energy consumptions (5%).
- PRI: 1,5 years
- The system can also be used to improve other systems like elevators, automatic stairs, HVAC or pumps in the airport
