A project designed to address the growing need for automated solutions in waste sorting, particularly in an aging workforce and a market with a high demand for sustainable operations. The project aims to leverage advanced AI and machine learning technologies to develop an automated system for detecting and sorting waste, particularly focusing on the identification of batteries in waste streams. This approach is not only more efficient but also enhances safety in recycling facilities.
TEMNOS focuses on addressing the complex challenge of recycling specific waste like vapes, laughing gas cartridges, and batteries. These items are problematic due to their mixed materials and hazardous contents, such as lithium-ion batteries in vapes that pose risks of fire or explosion and are considered hazardous waste.
The growth of vaping has significantly added to the electronic waste problem, with millions of these products discarded yearly, creating environmental hazards due to improper disposal and low recycling rates. The situation is worsened by the disposable nature of many vapes, leading to vast amounts of waste that are difficult to recycle and often end up in landfills or polluting waterways.
This highlights the urgent need for innovative recycling solutions to tackle these difficult-to-recycle items and reduce their environmental impact.
TEMNOS stands at the forefront of integrating sophisticated image recognition and X-ray technologies for precise waste identification. However, the project faces challenges, including the waste market's aversion to high-risk investments and a conservative approach towards adopting new technologies. Our strategy involves demonstrating the economic viability and superior performance of TEMNOS to overcome these hurdles. Additionally, we aim to reduce insurance costs for recycling facilities by mitigating the risk of fires, a significant concern in the industry.
TEMNOS operates in a competitive landscape where technological advancements by competitors like Fraunhofer-Gesellschaft present challenges. However, our strengths lie in our ability to quickly develop innovative technology, our expertise in AI, and our adaptability to a wide range of industries. Despite challenges like the need for significant investment and the lack of a full-scale prototype, opportunities abound in the growing EU waste recycling market and the increasing demand for advanced waste management technologies.
The project comprises three distinct components: Cyclops, Metes, and Argos, each tailored to specific sorting challenges and contributing uniquely to the waste management process.
Cyclops is like the smart helper in the world of waste sorting, using a camera and a scale to figure out what kind of trash is what. It’s really good at telling different types of waste apart, thanks to its smart tech. By teaming up with companies like STATICE BV, Cyclops shows its chops in making sorting tasks partly automatic, making things faster and more on point.
Metes is all about keeping things moving smoothly on the waste sorting line. It has this conveyor belt setup decked out with fancy sensors, including cameras and sensors that see stuff our eyes can't. It's working with Van Peperzeel BV to get better at picking out batteries from the bunch, aiming to sort things out quicker and smarter.
Argos is the heavy lifter for dealing with big piles of recycling stuff, especially good at spotting batteries hidden in shredded materials thanks to its X-ray vision. This is super important for keeping things safe by making sure batteries don’t end up where they shouldn’t. It’s a big deal for large recycling operations looking to avoid trouble and keep the recycling process smooth.
The strength of Temnos is its adaptability. Our expertise combined with our ability to analyze different waste streams, makes it possible to apply Temnos according to the wishes our clients.
We take the adaptability to the next level with the use of synthetic data. Currently in active development, a platform is being created to generate x-ray images of all kinds of waste streams, based on limited samples. This way the data collection process is significantly shorter, allowing us to create the most robust possible system based on your needs.
PARTNERS OF THE PROJECT
"GLIS provided us with the x-ray that we use for data collection and sampling, this way we can train the AI faster and more adaptive"