Thanks to its own funds, the Novika team is able to carry out internal projects aimed at developing new innovative tools or at becoming familiar with cutting-edge technologies and methods. This project is one of these innovative internal projects.
Novika works with multiple clients to produce laser textured surfaces. This internal project was intended to facilitate future projects. In the pulsed laser texturing process, seven experimental parameters can be controlled to modify the depth and roughness produced by the laser-matter interaction. In addition, the choice of these seven parameters also determines the production lead time. How can we predict the best combination of our seven adjustable parameters to achieve simultaneously the desired depth, lowest roughness and shortest laser processing time? This project aimed to simplify the process of finding optimal parameters.
We chose to involve artificial intelligence (AI) models in data analysis. We first carried out test shots to create a training database for AI. We then questioned these models in order to know the depths, roughnesses and production lead times for very large numbers of combinations. This mass of predictions was then filtered to match a user-targeted depth. The parameters associated with the lowest roughness and shortest time are then proposed for future projects.
