A leading airline service provider had developed an airline operator platform in reaction to the growth of customer in-flight sensor data. However, the company had a lack of good outlier detection models available meaning that, while each type of outlier detection problem was solved individually, there was no central service. And so Eraneos teamed up with the airline to try and fix these problems.
We got to work by implementing state-of-the-art anomaly detection algorithms. These were deployed as a service that could be reused in several applications inside the portal. We also deployed these in OpenShift to allow for continuous integration and testing.
Implementing these tools allowed the airline to make use of anomaly detection trigger warnings for customer technicians and ground crew. This also meant technicians could validate the anomalies thus optimizing the maintenance schedule. On top of this, we developed a high-availability Representational state transfer (REST) service for the company’s several internal applications.
As the project came to an end, we implemented production-level state-of-the-art outlier detection algorithms, established world-class data-driven and AI-powered operations platforms, and increased safety and reduced maintenance costs for the airline’s operators.
Our efforts had also helped the airline to increase its turnover within maintenance, repair, and overhaul (MRO).