Heavy Equipment|Case Study

Hydroelectric plants leveraged Cerebra Diagnostics for anomaly detection of a hydro generator to reduce downtime

The Client

One-Stop shop for all Hydro-electric power solutions based out of Asia playing the role of OEM, EPC and MRO.

Several thousand installation of high-performance turbines and generators across the globe.

The Problem

The generator unit in a hydro power plant is prone to different types of failures during the startup phase.

It takes a few days for a maintenance expert to perform root cause analysis and take corrective action

If operator ignores the failure and restarts the generator, then it results in a catastrophic failure in the long run which can result in an average downtime of three months

The client needed an early warning system that could detect anomalies in the running of the generator

Flutura's approach

Cerebra Diagnostics module was deployed to build an Early warning system for the client.
Streaming data from generators was recorded every second and included parameters like Active Power Generated, Generator Cooling Water Temperature and Generator Bearing Temperatures
Complex techniques such as Agglomerative Hierarchical Clustering and Neural Network with Multi-Layer Perceptron were configured and run on streaming data to identify early warning signals.


Generators across the globe with a capacity of 78,000+ MVA. Conservative calculation shows a saving potential of several hundreds of million USD


Prediction accuracy of identifying anomalies


Features given as input to complex algorithms, which helped identify deviation of 25x in a generator startup sequence in comparison to normal sequence