In the manufacturing process, there is a great need for more sophisticated data analysis – a need that has been brought to us by several players from industry. While showing simple deep learning inference models run on a smartphone, we identified great potential for parameter optimisation on CNC mills, better selection of rejects of 3D-printed parts, or in turning machines. Similarly, lots of data must be processed in space, spaceship components tested on the ground before launch, and/or collected data analysed on the ground. This can be automated using AI-based data analysis of all types of sensors. We are currently in the prototyping phase for a drill chuck for a CNC milling machine, or similar, in order to train the networks and set up the generic data platform. In a second step, MAS-Tech aims to transfer the approach of CNC sensor analysis to telemetry data from ESA’s Gaia, which (to our knowledge) is processed manually, and similar space application scenarios. We will then address customers. The total available AI-based sensor market, and also the serviceable obtainable market, is tremendous and growing rapidly, especially in Germany with the many mid-sized engineering companies.