Contactless torque sensor with AI software

Scope of application

The aim of the project is to develop a torque sensor for rotating shafts which uses artificial intelligence to detect inefficiencies and suggest preventive maintenance measures – thereby increasing the service life of the sensor and the systems it monitors.

How it works

Unlike conventional systems, which transmit measurement signals via electrical sliding contacts, the project solution utilises contactless signal transmission from the rotating shaft – without the associated wear and maintenance requirements of conventional systems. The sensor data is recorded and transmitted in real time, and evaluated directly on a processor integrated into the sensor using AI models. Unlike other contactless methods on the market, the solution incorporates AI-based data processing, which enables detailed process analysis and provides specific indications of inefficiencies and preventive maintenance measures.

Technical challenges

• Interference-resistant electromagnetic signal transmission from the rotating shaft
• Development of space- and energy-efficient electronics for integration into the sensor
• Design and training of AI models for predictive maintenance
• Optimisation of AI models for execution on embedded systems with limited computing capacity and low power consumption

Procedure

The project encompasses the mechanical design of the sensor, the development of the measurement electronics, modelling and training the AI using simulated and measured data, as well as comprehensive module and system tests.