VEGA Grieshaber KG
Sensor calibration using AI
How AI makes production more efficient? We have demonstrated this to sensor manufacturer VEGA Grieshaber in the field of sensor calibration with the help of machine learning.
The sensor manufacturer and world market leader VEGA Grieshaber KG from the Black Forest develops innovative measuring techniques that are easy to use and offer maximum safety and reliability.
Acceleration without loss of quality
VEGA wants to improve its production for the first time with the help of artificial intelligence. To this end, the calibration process for their sensors is to be optimized through the use of machine learning. The aim is to speed up the calibration process without any loss of quality.
Development of a machine learning algorithm
The AI engineers at mmmake have developed a machine learning algorithm that makes it possible to optimize sensor calibration and save up to 70% of the time required.
A precise two-stage process was used: In the first step, a comprehensive model was extracted that captures the common features of all sensors. In the second step, individual sensors were calibrated to take specific properties into account.
This approach enables customized calibrations for VEGA’s individual customer requirements.
Benefits
Data Engineering
Development of customized solutions that improve the data infrastructure.
Data Science
Generating comprehensive insights and added value through the use of advanced analytical techniques.
MLOps
Industrialization of AI projects through continuous model improvements and efficient monitoring.
We live technology! Our experts train in a wide variety of areas and are not afraid of new challenges posed by new technologies. Always up to date to offer you the best solution.
In our project with Vega, we relied on neural networks and the development of an individual machine learning algorithm with Python as the programming language. The powerful Tensorflow platform and joblib were used for modeling, which we used for multithreading in order to process the computationally intensive tasks in parallel and thus greatly reduce the overall execution time.