Bottle, Can or Coffee Cup?

By Dr Rob Gibbs, Materials Made Smarter Research Centre, Swansea University

How Computer Vision and Machine Learning can be used to Recognise Different Materials to Make Recycling Easier

This project has been developed by the Materials Made Smarter Centre at Swansea University in collaboration with the SUSTAIN Manufacturing Research Hub and Discover Materials to demonstrate how Computer Vision and Machine Learning can be used to recognise different objects to help with the sorting of materials for recycling.

The platform this project is built on is the Seeed Studio reComputer J1010 NVIDIA Jetson Nano 2GB Platform with the Arm Cortex A57 CPU and NVIDIA Maxwell GPU and it has been developed by Dr R. Gibbs and Prof. C. Giannetti based upon the NVIDIA DLI "Getting Started with AI on Jetson Nano” course.

The project has been successfully deployed on the Discover Materials stand at the Festival of Tomorrow in Swindon, 21st and 22nd February 2025. The event was great fun and over the two days more than 850 visitors came to the stand to learn about materials science, separation, characterisation and testing of materials. The demonstration worked at many levels and fit well with the other activities on the stand. Early primary school age children were excited to place different objects in the photobooth and see what the machine recognised them as, older secondary school age learners appreciated that the project was built on a small readily available computer using Python code that they themselves were familiar with and could achieve, and the adults appreciated the recent advances in smart sensor technologies and the real-world industrial application.

Several teachers attended the event and were excited about how they could bring the work to the classroom. If it is possible, future work could include porting the demonstration to a pack that could be downloaded onto existing school computers to teach about computer vision. The demonstration is currently heavily tied to the Jetson hardware and Jetpack environment so this is not a trivial aim, but worthwhile.

Overall, the demonstration prompted conversations with all visitors about the real world applications of computer vision to industrial sorting and about the machine learning specific concepts of the probabilistic nature of categorisation, reaction to new data, transfer-learning, re-training and feature extraction.

Interested in finding out more? Visit the Discover Materials website for all the documentation supporting the demonstration of this project.

Professor C. Giannetti would like to acknowledge the support of the EPSRC (EP/V061798/1) in this Materials Made Smarter Project.