General Electric is a major industrial player in the energy, healthcare and aeronautics sectors. Its GE Renewable Hydro division develops predictive monitoring and Industrial IoT solutions that include system behavior analysis. In this context, GE wants to implement a predictive maintenance solution for the inlet valves of its hydroelectric dams. A consortium was therefore created, in which CATIE is responsible for the work package on data acquisition and anomaly detection when these valves open and close.
A hydroelectric plant produces electricity from the potential energy stored upstream of a dam. Water is directed to the turbine through a penstock, and the inlet valve is a safety device that opens and closes the penstock. The turbine is controlled by a second, dynamically controlled set of gates to vary the plant’s output. At the request of GE Hydro and as part of the European DigiFed project, CATIE is developing a smart system to monitor the valve’s operation. The aim is not only to detect anomalies but also to predict future malfunctions using AI embedded in the system.
CATIE & GE Hydro are working together to design a solution to this problem. GE Hydro’s domain expertise made it possible to identify a set of sensors suited to detecting and identifying anomalies on the valve.
The acquisition system must collect the data and also interpret it to identify the state of the system. Data volume and the need for local processing are challenges for our embedded systems engineers and data scientists. The resulting prototype serves as a proof of concept, a prerequisite for the industrial prototyping phase.
This phase will be carried out by AGUILA technologies, which CATIE brought into the consortium so that the Nouvelle-Aquitaine ecosystem benefits from such a European project.
CATIE builds on its open-source platform 6TRON and can extend it with the new building blocks the project needs. The flexibility and modularity of this approach make it possible to adapt the solution to current needs and future challenges. GE Hydro is already considering new versions of the solution for other use cases. This project shows how complementary partners can take on the most complex challenges together.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 761708.
Technical data:
- Predictive maintenance
- Industrial monitoring
- Embedded AI
For more information: scp@catie.fr