Smart automation with Bossard and FHNW
The goal of the project is to push the limits of the unique joining technology for multi-material connections by developing a robotic-based automation approach coupled with a so-called smart solution module based on machine learning methods.
The Smart Automation project is bringing together 2 key aspects of the MM-Welding technology:
- Automation potential
- Big Data usage potential
The goal of the project is to push the limits of the unique joining technology for multi-material connections by developing a robotic-based automation approach coupled with a so-called smart solution module based on machine learning methods.
Many data from various sensors and transducers can be acquired during the ultrasonic process, that gives insights about the quality and performance of the connection. The Smart Solutions module developed in this project uses Machine Learning to provide a prediction of the connection quality and will be able to optimize the process parameters in an automated way.
Facts & figures
Consortium:
- MultiMaterial-Welding
- Bossard
- University of Applied Sciences Northwestern Switzerland
Funding:
Innosuisse
Budget:
1.19 Mio
Whitepapers:
- Realtime quality control of joining process, Bossard White papers, 2020, available online https://media.bossard.com/ie-en/-/media/bossard-group/website/documents/white-paper/bossard_whitepaper_smartsolution_en_07-2020_fv.pdf?la=en-ie
- From sandwich panels to fibrous materials; how to join lightweight materials with ultrasounds, AG Textilen Techniken, 2021, Online conference
- Experimental and numerical research of the MM-Welding joining technology for sandwich panels, Poster presentation, ITHEC 2020, Online conference
- Einblick in MultiMaterial-Welding und indie Aufgabenstellung, Veranstaltung für Mitglieder der Aargauische Industrie- und Handelskammer AIHK, 2021, Online conference
- MultiMaterial-Welding –Digitalisierung verbindet, CU Thementag «Digitalisierte Fertigung für eine emissionsfreie Mobilität», 2021, Online conference
Highlighted results
Robotic solution
A robot was equipped with a custom made end effector containing a servo driven ultrasonic actuator device and various sensors allowing acquisition of process parameters.
Several MM-Welding technologies have shown to be suitable for automated robotic assembly.
Machine learning method
Data acquired during the process are used to train machine learning algorithms and enable self-learning of the process to maximize connection strength and reliability.
Want to collaborate with us?
We are always looking for new partners to work on R&D projects.
We’d be happy to talk about it.
