1. Context

More than ever, the challenges facing the surface treatment sector are to improve process quality, reduce costs and lead times, and comply with environmental standards. This industrial reality is strongly impacted by numerous parameters, specific to all three players, which must not only be identified, but also controlled as effectively as possible. The answers to this multi-parametric problem cannot be obtained by the intuitive method, known as OVAT (one variable at a time), which consists of setting the levels of all the parameters except one, and measuring the desired response for several of its values. This tedious, time-consuming approach results in a wealth of information, often difficult to exploit, in which only the effect of each parameter is determined, without identifying possible interactions between factors.

Nowadays, when the time available for research...