Resource Use

AI waste sorting robot shortlisted for Young Inventors prize

The European Patent Office (EPO) has shortlisted engineers Victor Dewulf and Peter Hedley for the 2022 Young Inventors prize for their waste recognition and robotic sorting technologies, driven by artificial intelligence (AI).

Recycleye Victor and PeterWhat began as a treadmill, camera and a pile of waste collected from bins in Poole has been developed by the two entrepreneurs into a waste sorting system and company ‘Recycleye’. They have been named as one of three finalists for the prize, created by the EPO to encourage the next generation of inventors.

The pair said the project aims to increase the ‘purity, and therefore the value, of recycled waste bales’. In turn, it is hoped that this will heighten the financial incentive to recycle.

The invention consists of two parts, a ‘computer vision system’ – Recylcleye Vision – that uses AI to identify different types of mixed waste items and a robotic arm that ‘moves on six aces to autonomously pick out valuable material from a recycling conveyor belt’ with mixed, low-value waste.

Discussing this process with Resource, Recycleye stated that Recycleye Vision is used to ‘automate quality assurance’, and its robotic arm to ‘automate the quality control picking cabin at the end of the sorting process’. Currently, the arm can sort all dry mixed recyclables from households, including: paper, cardboard, cartons, aluminium, PET, HDPE, films etc.

The engineers also mentioned that the Recycleye Robotics system cannot work without Recycleye Vision, but the latter can be used alone. Installing the company's robotics includes the vision system, which identifies items and communicates the co-ordinates for the robotic arm to pick from and where to sort to. However, Recycle Vision can be deployed alone, for example, to provide a process with automated quality assurance.

According to Dewulf and Hedley, using AI to automate the process of sorting waste holds the potential to increase the proportion that is recycled by separating plastic and valuable waste from low-value, mixed waste. Its goal is also to avoid the more ‘expensive’, manual process of waste separation.

The computer vision system involved in the two-part invention uses a ‘cellphone-quality camera mounted above waste conveyor belts’, which reportedly takes ‘60 photos of passing waste per second’. These photos are then sent to an algorithm which ranks them for picking priority, before sending this information to the Recycleye Robotics sorting arm. The arm is then told where to pick up and place waste.

Together, the technologies can make ‘55 successful picks from a conveyor belt every minute’, says Recycleye.

Recycleye RoboticsSpeaking on the technology, Peter Hedley said: “Our visual recognition system can run on the fastest belts within a waste plant, which our competitors can’t do.

“The AI prioritisation of waste picking helps us get to about a 300 per cent performance increase – and a 300 per cent increase in the bottom line for our facility really helps them and their margins.”

Reviewing this increase, Recycleye highlighted the AI’s ability to pick high value items from the belt, in comparison to a manual picker selecting items randomly. Material Recovery Facilities, as reported by the company, can therefore “experience a 300 per cent increase in the value of their output.”

Recycle Robotics is already on the market, with installations at the FCC-managed re3 facility in Reading, England, Bryson Recycling in Northern Ireland, a further three units in MRFs across France and many more scheduled for the rest of the year.

Belgium-born Dweulf and British Hedley met while studying at Imperial College London. During his Master’s course in environmental engineering, Victor Dewulf visited a recycling  facility and noticed ‘how labour-intensive the waste sorting process was’.

This realisation, along with Hedley’s Master’s course in computer science, motivated Dewulf’s thesis on waste sorting automation using computer vision – an idea that attracted attention in the digitalisation-focused market. In 2019, he recruited Hedley to develop a prototype computer vision-powered waste recognition system.

The EPO prize recognises young innovators aged 30 or under who have ‘developed solutions to tackle global problems and help reach the United Nations’ Sustainable Development Goals. The winners of the Young Inventors prize will be announced at the European Inventor Award virtual ceremony on 21 June.

Antonio Campinos, EPO President, said: “With their twin waste recognition and sorting solutions, Victor Dewulf and Peter Hedley are making a vital contribution to reducing the world’s waste and moving towards a circular economy.

“The speed at which they have not just developed these innovations, but also turned them into reality, is remarkable and we look forward to seeing their story unfold.”