AutoWEEEdakt II - Automatic AI based WEEE recognition system

Purpose and goal

In the previous PiiA-funded project AutoWEEEdakt, we proposed and developed new innovative methods to create artificial intelligent (AI)-powered computer vision methods that could learn the categorization of WEEE by observing the manual sorting work being done. In this proposed project, we want to continue to increase the amount of labelled image data. Furthermore, we want to further develop methods that, through web scraping and the production of synthetic image data, improve the balance between the different classes and make our AI models more robust.

Expected effects and result

We expect that the improvements will enable consumer electronics manufacturers to more easily recover and reuse parts and components from their end-of-life products. With the innovations developed in AutoWEEEdakt and in the new AutoWEEEdakt II project, Sweden has the opportunity to strengthen its position as a forerunners in the recycling sector and demonstrate the country´s ability to keep up with the global development of applied AI solutions.

Planned approach and implementation

The project is led by Stiftelsen Chalmers Industriteknik, a research institute that works to develop innovative solutions for a sustainable future and has broad knowledge in applied AI and circular economy. The consortium includes Trapper Data, a consumer electronics manufacturer; El-Kretsen, a producer responsibility organization; NG Metall AB, an electronics recycler and OP technology an indtech company.