Dr. Völkel, what can the intelligent wearable do?

During manual assembly processes for electrical or fluidic connectors, it recognizes whether the assembly has been carried out correctly or not. We also count the plug connections so that no plug is forgotten. The wearable integrates seamlessly into the assembly process to perform this task so that the workers can continue working as before. The device learns continuously over its service life and receives regular updates.

What was the problem with previous solutions?

Some approaches to detection are based solely on detecting the click sound. However, this is not always sufficient because there are similar noises in production, which are then incorrectly perceived as the plug clicking. In addition, our choice of sensor technology and data processing enabled us to achieve a battery life of over eleven hours in production use without a large setup with a heavy battery.

How does the sensor glove work technically in detail?

The glove recognizes the movement of the worker and the sound of engagement, and only when both match the pattern that was previously trained is the insertion recognized as correct. The signals are still processed in the glove and both visual and haptic feedback is returned to the worker.

Dr. Anne Völkel: "The sensor glove constantly learns over the period of use and continuously improves thanks to AI."

© Hanna Bernardy

What role does artificial intelligence play in this?

We use an AI to explain to the device which plugs are properly connected, i.e. 'okay', and which are faulty, i.e. 'not okay'. This means that the AI evaluates the sensor signals in real time and transmits the feedback to the worker. In order to always be able to offer the best AI performance, the device learns with every connection in productive use and continuously improves.

Are there other areas in which the sensor glove can or should be used in the future?

We are currently testing the first stationary applications in which the adaptivity of our AI-sensor combination is being investigated for other tracking processes in production. Our vision is the general recognition of correct production processes, for the validation of which the combination of compact sensor technology and AI can be used.

Hall 10.0, Stand 530