Innovation of material quality control with artificial intelligence

AI powered material analysis

MiviA LogoThe Team of the projekt "MiViA" has developed an autonomous, self-learning microstructure analysis that combines the expertise of a materials scientist with the ability to recognize patterns using AI algorithms. The "MiViA" project started at the Institute of Metal Forming at TU Bergakademie Freiberg in October 2021. The MiViA team consists of four motivated scientific employees who are working on the development of a market-ready analysis software as well as the associated economic aspects.

On the way from a funded university project to an independent company, the start-up project "MiViA" is financed by the Federal Ministry of Economics and Climate Protection (BMWK) via an EXIST research transfer.

“MiViA” – Team: (f.l.t.r.) Roman Rostami, Miriam Corcoran, Grzegorz Korpala, Jessica Schneider

Automation of microstructure analysis

In the production and further processing of metallic semi-finished products, it is essential to determine the properties of the material so that the product can be proven to have the necessary quality for the respective application.
In this way, defects can be detected at an early stage and additional costs due to rejects and production downtimes can be avoided. In addition to various material tests, an analysis of the microstructure also takes place.
Currently, these analyses are carried out manually. This requires a qualified expert to obtain meaningful results. Due to a subjective evaluation of the expert, there is a higher result dispersion of the analyses and consequently a higher error rate.
The business idea of "MiViA" is therefore based on the commercialization of a self-learning, autonomous microstructure analysis system. Our software is based on an AI network structure trained with a database of 2 million images.
With the use of MiViA, metal processing companies have the possibility to make the analysis results reproducible and stable, independent of experts at any time. Thus, a cross-company comparison of the results is conceivable.


5 reasons for the use of "MiViA

5 reasons for the use of "MiViA

What makes MiViA special?
  • Autonomous image processing and analysis by AI engine
  • Database generation with unique MiViA process
  • Analysis independent of image quality
  • Reproducible, stable quantitative analysis results, independent of experts
  • From prototype to market-ready product
In order to adapt MiViA's software directly to the needs of industry and science, close cooperation with customers is important. For this purpose, feedback is obtained within an MVP (Minimum Viable Product) program. This means that the currently existing version of the microstructure analysis software is tested on the first potential customers. Comments can thus flow directly into the development and thus optimally adapt the software.
If you would like to test the capabilities of our software yourself, please contact jessica [dot] schneideratmivia [dot] ai. All further information will be communicated to you in the following.
Would you like to learn more about MiViA? Then subscribe to our newsletter. Sign up at infoatmivia [dot] ai with the keyword "Newsletter".

Gefördert durch: