This repository contains the source code of the paper: Zocco, F. and Malvezzi, M., 2025. CIRO7.2: A material network with circularity of -7.2 and reinforcement-learning-controlled robotic disassembler. arXiv preprint arXiv:2506.11748.
This repository contains the Python notebooks generating the results reported in the paper. Please note that only the notebooks for the case of "2 Parts and 1 Target" (abbreviated as "2P1T") contain, at the end, a visual simulation of the performance. The other notebooks do not show the video because, if included, they would have exceeded the file size limit allowed by GitHub (i.e., 25 MB). In those cases, the videos are available upon request.
IMPORTANT: GitHub may not be able to render the .ipynb files when you click on them. To visualize the .ipynb files without issues, download and upload them to Google Colaboratory. To download the repository, click on "Code" (the green button on the top-right) and scroll-down to find "Download ZIP".
Further details are provided in the preamble of each file.