In the PADAS project, we are working with Convadis AG to develop an AI-based method for the automated detection of minor vehicle damages, such as those typically caused during parking or low-speed maneuvers. The approach is based on acceleration data collected by GPS and telematics modules installed in fleet vehicles.

The goal is to develop a machine learning model that reliably distinguishes actual damage events from harmless movements. While existing systems are capable of flagging anomalies, they often generate a high number of false alarms. The focus is on detecting small to medium damages, especially those that remain unreported by users.

The main challenge lies in data-driven modeling using a limited amount of labeled event data.