Anomaly Detection

Detecting anomalies in a physical or industrial system, like a car battery, a wind turbine, or an antenna, is a crucial challenge. This allows for a quick investigation and/or repair of the system.

​

Adagos has developed several approaches to address anomaly detection problems:

​​

•   Creating a digital twin based on the system’s nominal behavior to detect any deviation from its normal operation.

•   Creating an autoencoder to compress the data characterizing the system's operation and identify anomalies from normal cases using clusters.​

Adagos has developed a parsimonious AI, oriented to technical data.

Our technology is specifically adapted to detect anomalies in physical systems.

​

Adagos has built a robust car battery model capable of detecting various anomalies, enabling predictive maintenance for the battery.