Periodic Reporting for period 1 - Grinner (ImprovinG waste management through an AI-powered detection system of batteries utilising data from X-Ray detectors and pick-and-place robots.)
Reporting period: 2022-09-01 to 2024-02-29
A key issue that is currently affecting the Waste Electrical and Electronic Equipment (WEEE) management chain is that of battery-caused fires, costing waste management facilities millions of euros every year and acting as a strong barrier to making Europe circular and carbon neutral. Some battery types located inside discarded WEEE, in particular lithium-ion (Li-ion) and nickel-metal hydride (NiMH), can ignite or explode when damaged (e.g. when entered an electronic scrap recycling shredder within a recycling facility).
Increasing occurrences of waste fires that are caused by improperly discarded, mainly Li-ion ones, threaten the whole waste management sector in numerous countries. Ιn recent years, high quantities of these batteries have been found in several municipal solid waste streams of many European countries, including Austria, Germany, France, and Sweden. Recent research showed that high amounts of lithium-ion batteries (LIBs) are improperly discarded in different municipal solid waste streams, such as residual household waste, lightweight packaging waste, or metal packaging waste. While there is a plethora of different electrochemical systems, the average distribution is shifting more and more towards metal lithium and lithium-ion. That shift is accompanied by increased fire hazards and other safety challenges all along the value chain in the batteries’ end-of-life.
AS IT STANDS, THERE IS NO AVAILABLE SOLUTION TO ERADICATE THE FIRES CAUSED BY THESE SO-CALLED 'ZOMBIE' BATTERIES.
The GRINNER project aims to commercialise an autonomous AI-enabled robotic sorting system capable of detecting and removing e-waste containing batteries from current waste streams before they enter machines that crush and consolidate waste, causing damage to batteries, and massively increasing the risk of fires. The system will comprise of the following:
-The fastest energy-resolved X-Ray detectors in the market;
- A Machine Learning-enabled software module that will analyse X-Ray data and effectively detect waste containing batteries while passing through the waste flow
The GRINNER project objectives are to:
- Build an X-Ray data set of e-waste;
- Develop an AI software module for detecting batteries in e-waste;
- Develop a prototype system and install it in an e-waste facility to conduct live trials;
- Explore the potential for exploiting GRINNER as an economically viable, stand-alone product for e-waste treatment plants. passing through the waste flow.
Increasing occurrences of waste fires that are caused by improperly discarded, mainly Li-ion ones, threaten the whole waste management sector in numerous countries. Ιn recent years, high quantities of these batteries have been found in several municipal solid waste streams of many European countries, including Austria, Germany, France, and Sweden. Recent research showed that high amounts of lithium-ion batteries (LIBs) are improperly discarded in different municipal solid waste streams, such as residual household waste, lightweight packaging waste, or metal packaging waste. While there is a plethora of different electrochemical systems, the average distribution is shifting more and more towards metal lithium and lithium-ion. That shift is accompanied by increased fire hazards and other safety challenges all along the value chain in the batteries’ end-of-life.
AS IT STANDS, THERE IS NO AVAILABLE SOLUTION TO ERADICATE THE FIRES CAUSED BY THESE SO-CALLED 'ZOMBIE' BATTERIES.
The GRINNER project aims to commercialise an autonomous AI-enabled robotic sorting system capable of detecting and removing e-waste containing batteries from current waste streams before they enter machines that crush and consolidate waste, causing damage to batteries, and massively increasing the risk of fires. The system will comprise of the following:
-The fastest energy-resolved X-Ray detectors in the market;
- A Machine Learning-enabled software module that will analyse X-Ray data and effectively detect waste containing batteries while passing through the waste flow
The GRINNER project objectives are to:
- Build an X-Ray data set of e-waste;
- Develop an AI software module for detecting batteries in e-waste;
- Develop a prototype system and install it in an e-waste facility to conduct live trials;
- Explore the potential for exploiting GRINNER as an economically viable, stand-alone product for e-waste treatment plants. passing through the waste flow.
The GRINNER project started in September 2022 and has made very good progress in the first reporting period. All deliverables and milestones have been achieved and making good progress towards set objectives. Within reporting period 1,
- The technical specifications and requirements of the GRINNER solution were finalised after an in depth exercise to ensure that all the required features are not omitted including regulatory which are essential to ensure the success of GRINNER within the markets place
- The early X-Ray prototype was deployment and work commenced for building of a database of WEEE scrap. The technical teams are making good progress and within year 2, aim to achieve the set project KPT of 1500 datasets.
- Automation sorting devices have been explored and the team identified suitable solutions for the sorting that are in development for implementation and integration within reporting period 2
- The system conveyor belt as well as the test shield and lead-lined curtains have been ordered in accordance with the system specifications for integration within reporting year 2 at end user site
- The technical team, led by the University of Essex progressed in the battery detection process where both synthetic and real data was used to train the AI model. A key outcome of these tests demonstrated that the synthetic data does have a positive impact on the accuracy of the AI model indicating good traction towards achiving a high accuracy in the detection
- The coordinator LYNQ also made good progress on the development of the Manufacturing Execution system where most of the back end algorithms have been developed and awaiting integration with the GRINNER subsystems for integration. Within reporting period 1, the consortium led by LYNQ established and agreed the system architecture and by the mid term began developing the application interfaces that will be used. In addition, the system user interfaces have been developed and a mock-up demonstration performed to the consortium.
Thanks to the rigorous work ethic of the GRINNER project partners, GRINNER is on track and has gained good momentum that will no doubt see the team advance the development in a timely manner in reporting period 2.
- The technical specifications and requirements of the GRINNER solution were finalised after an in depth exercise to ensure that all the required features are not omitted including regulatory which are essential to ensure the success of GRINNER within the markets place
- The early X-Ray prototype was deployment and work commenced for building of a database of WEEE scrap. The technical teams are making good progress and within year 2, aim to achieve the set project KPT of 1500 datasets.
- Automation sorting devices have been explored and the team identified suitable solutions for the sorting that are in development for implementation and integration within reporting period 2
- The system conveyor belt as well as the test shield and lead-lined curtains have been ordered in accordance with the system specifications for integration within reporting year 2 at end user site
- The technical team, led by the University of Essex progressed in the battery detection process where both synthetic and real data was used to train the AI model. A key outcome of these tests demonstrated that the synthetic data does have a positive impact on the accuracy of the AI model indicating good traction towards achiving a high accuracy in the detection
- The coordinator LYNQ also made good progress on the development of the Manufacturing Execution system where most of the back end algorithms have been developed and awaiting integration with the GRINNER subsystems for integration. Within reporting period 1, the consortium led by LYNQ established and agreed the system architecture and by the mid term began developing the application interfaces that will be used. In addition, the system user interfaces have been developed and a mock-up demonstration performed to the consortium.
Thanks to the rigorous work ethic of the GRINNER project partners, GRINNER is on track and has gained good momentum that will no doubt see the team advance the development in a timely manner in reporting period 2.
The GRINNER project is halfway into the development time and within reporting period 1 has achieved the following progress towards set impacts:
- demonstrated progresses towards effectively identifying batteries within WEEE waste
- identified an effective sorting method that can be used to sort the WEEE waste
- produced a proof of concept multi-energy photon counting detector for the X-Ray system
- Produced a proof of concept x-ray system capable of detecting batteries within WEEE waste
Reporting period 2 will be focused on the implementation of the system, integration and validation trials. Extensive testing will be performed demonstrate the full set of impacts set. Specifically:
- An effective method to eliminate fires associated with batteries
- A stand alone GRINNER system that is cost effective with demonstrable ROI
- Reduction of staff in hazardous environment
- A system that will reduce the environmental impact of electrical waste
- A system that will enable proper handling of WEEE waste
- demonstrated progresses towards effectively identifying batteries within WEEE waste
- identified an effective sorting method that can be used to sort the WEEE waste
- produced a proof of concept multi-energy photon counting detector for the X-Ray system
- Produced a proof of concept x-ray system capable of detecting batteries within WEEE waste
Reporting period 2 will be focused on the implementation of the system, integration and validation trials. Extensive testing will be performed demonstrate the full set of impacts set. Specifically:
- An effective method to eliminate fires associated with batteries
- A stand alone GRINNER system that is cost effective with demonstrable ROI
- Reduction of staff in hazardous environment
- A system that will reduce the environmental impact of electrical waste
- A system that will enable proper handling of WEEE waste