With 300 new applications per year, Constellium Valais produces a wide variety of aluminum objects using the extrusion process. Each new request requires significant engineering effort, particularly to determine if the desired profile can be manufactured and at what cost, and to design the tool and production recipe. One of the essential tasks of this analysis is to search for profiles similar to the requested one in the production history to complete the personal experience of specialists who study its manufacturability and predict the pressure of the press during extrusion.
The P3 (Press Pressure Prediction) Innosuisse project, coordinated by the SPL research group of the HEI Valais and in partnership with Constellium Valais SA, aimed to use AI technologies to help experts conduct the feasibility study. This goal was achieved through the development of powerful tools to search for similar profiles in the production history as well as press pressure prediction. At Idiap, the Perception & Activity Understanding group participated in this collaboration.
How it works
To obtain the necessary data for AI methods, Constellium made significant efforts to retrieve all production data, including over 1,200 drafts of already manufactured profiles and data from 730,000 extrusions of these profiles. This information was treated to design and train two models: a neural network that extracts a characteristic signature of the shape of a profile from its image and quickly finds similar profiles in the database, and a conditional neural network that predicts the pressure during extrusion for a specific speed and temperature.
Finally, a graphical interface developed by HEI was implemented to allow engineers and operators to use the models easily according to their daily needs.
The experts of Constellium Valais are very satisfied with this new tool, as it allowed them to quickly find profiles similar to the one to be evaluated and provided them with new means of analysis to quickly simulate pressure throughout the extrusion process. This project is a fruitful collaboration between research and industry in Valais, and the partners at Constellium have found new ways to use and improve model predictions.
About the Idiap
The Idiap Research Institute has made significant contributions to the field of AI and to the canton of Valais. Its groundbreaking research has helped to shape the development of AI technologies, and its commitment to community outreach and education has helped to foster innovation and growth in the region. As the field of AI continues to evolve, Idiap is sure to remain at the forefront of research and innovation, paving the way for the next generation of AI technologies.
