Purpose: Reducing the variable portion by predicting faults in substation equipment using artificial intelligence.
Objectives
Development of computer programs for highly accurate numerical methods, in order to provide legitimacy and savings in the acquisition of land and installation of wind turbines.
Key Results
- Technical: Establish a protocol for monitoring and continuously improving the AI solution, reviewing and adjusting the algorithms every quarter, based on the results obtained and lessons learned
- Strategic: Increase the accuracy of fault predictions in transmission substation equipment by 15% within one year, improving the efficiency of the predictive maintenance process
- Economic: Reduce unplanned downtime in transmission substation equipment by 10% within one year
Main information
Scope
Development of a computational platform with Machine Learning models based on data from operation and maintenance systems for transmission assets at Furnas substations
Partners / Entities of support
PUC-RJ
Investment (contract)
R$ 12MM
Scalable?
Yes
Stages
Design
May/2020
Kickoff
December/2020
Prototype
April/2023
Tests
July/2023
Implementation
November/2023
Technology Transfer
December/2023
Deadline for implementation (in months)
36 months
P.S.:
Deadline: December/2023
