Data-based model applied to thermoforming process control

Thermoforming, a pivotal technique in the manufacturing of thermoplastic products, has traditionally relied on operator experience for process control. In this work, we use integrated sensors into a thermoforming machine to gather data using a design of experiments that feed a statistical prediction model. A database with sensors data and corresponding geometric quality measurements has been built that simplifies the development of the statistical model and its deployment online during production to monitor and control the process. This process control is automatic and much more reliable than human means. A prototype test case is related and proves the validity of our method that is ready to be tried on a more realistic industrial test case.