This project is funded by the FFG as part of the “AI for Green” program.

Voestalpine High Performance Metals DIGITAL SOLUTIONS GmbH is part of Voestalpine High Performance Metals the global market leader for tool steel and other special steels and part of the Voestalpine Public limited company. Voestalpine High Performance DIGITAL SOLUTIONS will contribute the industrial expertise and domain knowledge in the field of industrial manufacturing processes and the use case to reach high sustainability goals. SAELING will build upon initial platform solutions for machine data acquisition and visualization. The envisaged solution is based on data collection from sensors over edge computing up to data visualization. Edge computing as a core capability is currently used for low performance data capturing based on industrial protocols e.g., OPC UA and Profinet. The open challenge is the heterogenous environment within manufacturing companies on a global scale.

Siemens is experienced in research as well as the technologies required for the production of goods. The company contributes a strong portfolio in the digitalization of production and automation systems. Siemens' participation in SAELING is carried out by its horizontal research and development unit with its research group 'Configuration Technologies', which supports the Siemens business units with their production processes of large and complex products, has profound expertise on computational tasks arising during configuration and production as well as the algorithms and heuristics necessary for solving them. Siemens strives for sustainability with new technologies and is fully committed to turn its own operations carbon-neutral by 2030 and to help its customers avoid hundreds of millions of tons of CO₂e emissions per year.

The University of Klagenfurt is the coordinator of SAELING. The project members from the University of Klagenfurt bring in extensive research and project experience on the design and application of Al methods. Their expertise on optimization methods for challenging practical applications covers symbolic Al paradigms such as Answer Set Programming, Boolean Satisfiability, Constraint Programming and Integer Linear Programming, as well as sub-symbolic Al approaches in Classification Methods, Genetic Algorithms and (Deep) Reinforcement Learning.

Prof. Guns of the KU Leuven in Belgium is a leading expert on integration prediction and optimization tasks (including an ERC Consolidator Grant on the topic of 'Conversational Human-aware Technologies for Optimisation'). More specifically his expertise of decision-focused learning, where downstream optimization problems can be used in the loss function, such that the machine learning training focuses on errors that impact the decision problem most.