Within the use case n.4, the new O&M strategy concept is data-driven and based on analytics tools able to provide an early prediction and detection of faults.

As pilot leader, BFP is making available eight utility-scale PV plants deployed in the south of Puglia region, in Italy; historical and real-time data at inverter level for each dataset are provided. A well-defined API enables Apache NiFi to retrieve all the necessary information; data are gathered for all eight datasets and in order to preserve as much relevant data as possible, the API that sends data is called. These data are integrated, sent to Kafka and persisted into a MongoDB collection.

The analysis of such information and the application of Artificial Intelligence will provide, as final result, an assessment about the health state of the assets and a proposal of predictive maintenance.

In terms of middle/long-term outcomes, improvements of operational efficiency, increased self-consumption and electricity cost reduction are expected.