Ansmann AG
Contact at the AI Progress Center
Marco Huber
AI potential
Initial situation
Ansmann AG produces solutions in the fields of rechargeable batteries, charging technology and power supply units. Production times vary depending on the general conditions and the composition of the production team. Planning quality can be improved by accurately predicting production times.
Furthermore, the later faults are detected in production, the more costly they are to rectify. One possible shortcoming is the incorrect assembly of the battery housings with the cells (plus and minus poles).
Solution idea
Several use cases for Ansmann AG were discussed as part of the AI Explorer. The following proposed solutions were discussed:
A) Using AI to predict production times and combining this with optimization to generate suggestions for personnel planning.
B) Detection of the installation direction of the battery cells using AI and detection of installation errors by comparison with a pattern.
Benefit
A) The benefit is the prediction of the production time required for a specific order and the generation of a proposal for the allocation of workstations for this order.
B) Optical detection of incorrect assembly means that either quality checks can be carried out more quickly or errors can be detected during assembly.
Implementation of the AI application
A) Collection of data from different team combinations including framework conditions and actual production time, training of the AI, definition of targets for the optimization of personnel planning, connection of the optimization to the AI.
B) Generation of annotated images of the battery (good and bad parts) with camera system, training of the AI to recognize the installation direction, determination of the target occupancy of the battery, comparison of the prediction by the AI and the pattern.