The RecAIcle Consortium met for a strategic meeting on July 10, 2024, at the Digital Waste Research Lab (DWRL) of the Montan University of Leoben (MUL) in St. Michael. RecAIcle aims at achieving improvements in manual sorting in recycling trash separation plants, through learning from and interactions with the sorting worker. This complex learning and interaction system requires the implementation of cutting-edge Artificial Intelligence (AI) and Computer Vision (CV) methods. Trash recycling, especially plastics, poses unique challenges to CV systems, due to the visual appearance of deformed plastics, and to the variations in trash composition over time.
The purpose of the meeting was to assess the progress made during the second year of the project and to establish a strategic plan for the next phase. Discussions focused on the coordination of experiments at the DWRL, particularly the acquisition of necessary waste samples, for example, recycling plastics.
In addition, the hardware and software infrastructure for the upcoming recAIcle prototype were discussed. This included updates on various algorithms, such as waste particle detection and classification and action recognition, as well as the integration of components across different teams. The event concluded with demonstrations at the DWRL (MUL) by Pro²Future and Siemens, which highlighted the practical applications derived from the consortium’s research.
The participants of the meeting were (from right to left in the picture): Reinhold Seiss (Siemens), Sebastian Schreiner (Siemens), Bernd Hödl (Siemens), Michael Haslgrübler (Pro²Future), Benedikt Häcker (Siemens), Johannes Schürz (Siemens), Julian Aberger (Montan University of Leoben), Martin Höffernig (Siemens), Michael Krisper (Pro²Future), Georgios Sopidis (Pro²Future), Michael Siegl (Pro²Future), Renato Sarc (Montan University of Leoben), Jesús Pestana (Pro²Future).
Article by Jesus Pestana