Impact

Key Marketable Innovations

The main key marketable innovation consists of two software components, respectively, one implementing the preference-based optimization algorithm and one implementing the hybrid optimization algorithm (i.e., combining qualitative and quantitative optimizations). Such software components will be provided with a GUI to enhance the interaction with the user during the process optimization procedure. In such a way, even non-expert programming users will be able to perform the optimization of a target process with advanced algorithms. In addition, the expertise of the users can be exploited in an easy and intuitive way for the optimization of the considered process.

Such software components will be developed to provide a general (i.e., not customized on specific processes) optimization tool that can be applied to any generic application. Indeed, the developed GUI will be flexible to make the user able to define the target optimization variables, quantitative metrics, and all the required quantities for process optimization purposes.

The two software components will be made available to customers to guide their process optimization procedures.

Target Groups

There are no specific target user groups. In fact, the developed software components will be general and applicable to any process/application requiring a static tuning of their parameters (e.g., machine tuning for maximizing output performance). In addition, the proposed software components will make non-expert programming users able to exploit advanced AI-based algorithms for process optimization purposes. This will be possible by means of the developed GUI, easing the usage of such algorithms (providing a natural way of interaction between the user and the methodologies).

For the two use cases, we will ensure that the solution is specific to our use case partners (TEC-EUROLAB and SMARTZAVOD), in order to gain the knowledge required to develop a generalized solution. This means however that these two use cases will have a Time to Market of 0, generating economic benefit immediately for these partners and their customers. The generalized solution will be targeted at SME manufacturers throughout Europe, but the initial phase will target those SMEs that are part of Santer and IIC network in Italy, Germany, Slovak Republic and beyond.

Requirements and potential barriers 

The main barrier related to the adoption of the proposed technologies is the level of digitalization of the target (possible) customers. While the developed software components will be independent of the target process/industry sector/company, the complete advantages will be achieved within a connected production plant. In fact, by connecting the software components directly with the target processes/machines it will be possible to automatically update their parameters from the software components during the execution of the optimization process. In such a way, most of the manual operations to be performed by the user will be removed, fully automatizing the optimization procedure. Successful integration of a quantitative optimization algorithm (i.e., connecting the optimization algorithm with the target process, the database, and the GUI) has been already provided by SUPSI in the EIT-M IMPALA project. In this project, Bayesian optimization has been employed to optimize two processes, fully connecting the optimization algorithm to the industrial process (i.e., automatizing the optimization procedure). Indeed, this know-how will be employed in this project in order to improve the usability of the proposed software components.

No barriers are highlighted w.r.t. the users (i.e., no specific programming expertise will be needed - a GUI will be designed to increase the usability of the optimization algorithms - and no other issues/limitations are found).