AI-powered Detection of Parking Damages Based on Acceleration Sensor Data (PADAS)

In the PADAS project, we collaborate with Convadis to develop an AI-based method for detecting minor parking damages using acceleration sensor data. The aim is to distinguish real damage from false alarms.

Key data

Description

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.