Anomaly Behavior Analysis for Sensors Fault Detection

In today's world, sensors play a crucial role, as they feed information to make accurate decisions and take actions; therefore, making sure that sensors behave correctly is critical. In this work, we focus on inspecting the data provided by sensors, aiming at discovering any issue due to malfunction, misuse, or any other source of error before the issue is propagated through the system. To achieve that, we propose a novel approach based on wavelets embedded in a microcontroller to analyze data from sensors. The objective is to determine whether the sensor is issuing correct data (normal behavior) or not (abnormal behavior), to prevent the error from reaching other parts of the system.