The RECLAIM Project has reached another milestone.
In their first review meeting, the project consortium partners presented their progress update to the European Commission panel.
The initial feedback from the European Commission reviewers was positive and they expressed satisfaction with our project’s overall progress. Specifically, they commended the project’s achievement in seamlessly integrating all systems within the portable material recovery facility (prMRF).
Furthermore, the reviewers expressed confidence in the project’s trajectory to achieve the objectives outlined in the Grant Agreement. Their positive feedback serves as a validation of the collective and reinforces our commitment to delivering impactful results.
Summary of work progress until M18
RECLAIM focuses on the development of a portable, robotic Material Recovery Facility (prMRF) to efficiently sort recyclable waste in small communities and less accessible regions, thus aiming to provide a cost-effective solution tailored to small-scale material recovery. The project exploits well-tested technology in robotics, computer vision and Artificial Intelligence (AI) which is improved further to achieve industrial level decentralised processing of waste.
READ how RECLAIM is contributing to EU’s Circular Economy goals
Following partners’ inputs on the prMRF’s operational context, a draft outline has been created to guide the technical efforts to develop the prMRF. For data collection, target materials were specified to understand their type, such as PET bottles, LDPE film, HDPE bottles, PP/PS, Ferrous, Aluminium cans, and Drinking cartons, and their “possible” compositions, following which RGB and HSI cameras were validated after lab tests and installed in the prMRF.
RECLAIM delivers simpler, cost-effective robotic solution Robotic Recycling Workers specifically designed for sorting waste
Then, experiments were conducted with complementary gripper mechanisms to understand if they can fit into waste treatment applications. To carry out sorting activities, a new robotic system, like the Robotic Recycling Workers (RoReWos), designed for this purpose, has been employed. This is purely due to its simplified hardware design and ease of installation in prMRFs to reduce implementation costs and increase material recovery efficiency with only minor reductions in performance and productivity.
Also read: RECLAIM’s AI-Driven Waste Sorting Robots Could Accelerate Circular Economy Transition
Additionally, a new computational approach was developed to analyse the coordination between novel grippers and RoReWos for certain waste recovery tasks.
Next phase will focus on improving picks/minute rate
In the next phase of the project, further improvements will be made regarding the targeted performance by, for instance, increasing the successful picks/minute rate.
As for the prMRF itself, consortium partners have successfully implemented the container that houses the mechanical equipment for waste processing and made adjustments to facilitate the installation and operation of robotic and AI components.
The successful implementation of the prMRF has enabled data collection to train the Artificial Intelligence models and support the Recycling Data Game (RDG). All of this was done in close coordination with other work packages to ensure proper fit of individual technologies with only small adjustments during integration. For example, ensuring the placement and coordination of mechanical equipment alongside the computer vision, HSI, and robotic hardware threw up practical challenges but were resolved with the partners’ technical support.
RECLAIM successfully evaluated integrated performance of fully assembled prMRF in realistic conditions
Finally, in relation to the Recycling Data Game (RDG), rich data sets were gathered after multiple rounds of data collection. Participatory design sessions with AI stakeholders and focus groups with environmental experts helped design themes and challenges for the RDG. As a result, seven mini games were created to test annotation feasibility on mobile devices. A prototype with annotation challenges has been tested, with plans to add content testing challenges and enhance user experience.