SUDE Piloot!

Innovative technology for automatic disassembly of electronics

In recent years, KU Leuven's Re- and Demanufacturing lab has developed new technologies to automatically dismantle waste electrical and electronic equipment (WEEE). Indeed, rising labour costs and the increasing use of glue and snap connections make it increasingly challenging to open discarded appliances in a cost-efficient way and sort the components for material recycling. 

In the SUDE Pilot! project, KU Leuven worked with machine builder Valvan and recycling company Galloo to further develop these new technologies and merge them into a pilot setup. The setup is designed to perform automatically the most labour-intensive task of processing discarded laptops: removing the screws. The main elements of the setup are a six-axis robot, a camera setup and a waterjet cutting machine. 

Since the screw coordinates vary for each unique laptop model, we have also developed the necessary software to distinguish different models from each other and add new, not yet recognised models to a database along with the identified screw coordinates. Thus, we will eventually create a database with disassembly instructions for each model.

MOST IMPORTANT
RESULTS

  1. We realised a first industrial set-up that enables automatic disassembly of WEEE. In doing so, we use AI, a six-axis robot and a waterjet cutting machine. 
  2. We designed an algorithm, YODO, that can distinguish different laptops from each other and add new models not yet recognised to a photo database. 
  3. We only need to process 2,700 laptops with the setup to achieve a 'retrieval ratio' of 87%. This means that from then on, the setup can cut out the screws of almost 9 out of 10 laptops in one cycle, drastically speeding up the process.

MOST IMPORTANT
LESSONS LEARNED

  1. The demand for end-of-life components for reuse has increased dramatically. This represents a great opportunity to further develop automatic disassembly techniques to make the process even more cost-efficient in the future. 
  2. Artificial intelligence will be increasingly integrated into industrial setups in the future. The platforms used in industrial setups are not yet adapted for this purpose, posing additional challenges. 
  3. The government could help remove barriers by, for example, allowing the recovery of WEEE after collection in container parks, reducing VAT on recovered products, subsidising further development of innovative systems ... 
  4. One of the biggest challenges for the Circular School is that the working group consists solely of volunteers. In terms of workload and coordination, this occasionally results in stress and last-minute work.
5.500 Photographed laptops
87% Recognised laptops

WHAT DOES
THE FUTURE HOLD?

We are currently working on a robotic screw removal system that uses the developed product identification software to make the whole system (self)learning. Screw removal with this system is slower compared to waterjet cutting and requires high precision. 

There are also challenges, such as removing rusted screws and snap connections, for which we are exploring the possibilities of human-machine collaboration.