📒 𝗡𝗲𝘄 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗽𝗮𝗽𝗲𝗿 𝗳𝗿𝗼𝗺 𝘁𝗵𝗲 𝗖𝗶𝗿𝗰𝘂𝗯𝗼𝘁 𝗣𝗿𝗼𝗷𝗲𝗰𝘁: Advancing robotic waste sorting with AI ♻️

The rise of LEAN and small-batch manufacturing calls for 𝗳𝗹𝗲𝘅𝗶𝗯𝗹𝗲 𝗮𝘂𝘁𝗼𝗺𝗮𝘁𝗲𝗱 𝘄𝗼𝗿𝗸𝘀𝘁𝗮𝘁𝗶𝗼𝗻𝘀 that can seamlessly switch between different waste types over time. Our latest study explored deep learning models from the Segment Anything Model (SAM) family for extracting waste objects from raw images, followed by a classification stage to ensure accurate waste sorting. This two-step procedure eliminates the need for dedicated waste detection and segmentation algorithms, enhancing robotic waste sorting by efficiently segmenting and classifying highly variable objects.

Tested on 𝗳𝗼𝘂𝗿 𝘂𝘀𝗲 𝗰𝗮𝘀𝗲𝘀 (floating waste, municipal waste, e-waste and smart bins), 𝗼𝘂𝗿 𝗺𝗲𝘁𝗵𝗼𝗱 𝗮𝗰𝗵𝗶𝗲𝘃𝗲𝗱 𝟴𝟲-𝟵𝟳% 𝗮𝗰𝗰𝘂𝗿𝗮𝗰𝘆, demonstrating its robustness and industrial applicability.

By simplifying deployment, improving efficiency, and reducing costs, this approach has the potential to enhence robotic waste sorting and boost recycling and material utilization in manufacturing.

You can explore the full study here
𝗔𝘂𝘁𝗵𝗼𝗿𝘀:

Arso Vukicevic, Milos Petrovic, Nebojša Jurišević, Djapan Marko PhD, Nikola Knežević, Dr Aleksandar Novakovic & Kosta Jovanovic

𝗦𝘁𝗮𝘆 𝘁𝘂𝗻𝗲𝗱 𝗳𝗼𝗿 𝗺𝗼𝗿𝗲 𝘂𝗽𝗱𝗮𝘁𝗲𝘀 𝗳𝗿𝗼𝗺 𝗖𝗶𝗿𝗰𝘂𝗯𝗼𝘁 𝗮𝘀 𝘄𝗲 𝗮𝗽𝗽𝗿𝗼𝗮𝗰𝗵 𝘁𝗵𝗲 𝗳𝗶𝗻𝗮𝗹 𝗽𝗵𝗮𝘀𝗲 𝗼𝗳 𝘁𝗵𝗲 𝗽𝗿𝗼𝗷𝗲𝗰𝘁! While the project is nearing its conclusion, this milestone brings exciting new insights, results, and key takeaways that we’ll be sharing in the coming period.

This research was supported by the Fond za nauku Republike Srbije – Science Fund of the Republic of Serbia, GREEN – CircuBot