Search

How AI Helps Recyclers Turn Mixed Waste Into Reusable Material | NVIDIA Blog

The article describes how AI and robotics can identify and sort recoverable materials from mixed waste. It introduces four NVIDIA Inception startups working to keep materials in circulation, including AIRECO, which focuses on plastic waste.

Recycling Waste with Microbes: A Path to Metal Extraction

Translated title. Original: Переработка отходов с помощью микробов: путь к добыче металлов

Rising metal demand is increasing pressure on mines, while electronic waste remains a largely untapped source of valuable materials. The authors call for pilot projects, shared investment risks, and common standards to move biological extraction methods from labs into use.

Attached caps are just the beginning. More changes are coming to the bottles - ddbelchatow.pl

Translated title. Original: Przyczepione zakrętki to dopiero początek. Idą kolejne zmiany w butelkach - ddbelchatow.pl

EU rules will gradually raise recycled plastic use in PET beverage bottles and keep caps attached to reduce waste; colors may vary with recycled content. Phases include 25% recycled content by 2025, 30% by 2030, and 65% by 2040, affecting bottle design and consumer experience.

Conference on recycling: critical minerals, batteries, and AI-related electrical and electronic waste - News and Statistics - IndexBox

Translated title. Original: Conferenza sul riciclo: minerali critici, batterie e rifiuti elettronici legati all'IA - Notizie e Statistiche - IndexBox

The conference bundles three parallel tracks at the Cobb Convention Center, focusing on AI-related electronic waste, battery recycling economics, and regulatory policy for critical minerals. Key talks address domestic nickel refining, North American supply chains, e-waste management, and sustainable design of gigafactories, with industry leaders from Westwin Elements, Basel Action Network, and Google Cloud contributing. The program links recycling, resource security, and policy with participation from technology firms, regulators, and recyclers, culminating in a session that ties together the event's overlapping themes.

AI transforms waste impurities into useful ingredients for a stronger recycled iron

Translated title. Original: L’IA transforme les impuretés des déchets en ingrédients utiles pour un fer recyclé plus fort

Global metal output is around 2 billion tonnes per year, driven by industry, infrastructure, and energy. Heavy reliance on primary materials drives significant emissions and energy use; increasing recycled metals could lessen environmental impact and reduce reliance on critical raw material imports.

A camera- and hand-equipped robot is revolutionizing waste management – up to 430,000 euros in additional revenue | Tivi

Translated title. Original: Näin kameroilla ja kourilla varustettu robotti mullistaa jätehuollon – Jopa 430 000 euron lisätuotto | Tivi

Scottish startup Danu Robotics develops AI-powered waste-sorting robots that can retrofit into existing facilities, aiming to cut manual labor and boost material recovery; they seek about USD 5 million in funding, with potential annual revenue gains around EUR 430,000. Similar systems exist, and Finnish pilots have explored AI in waste recognition.

Recycler: Cost-Free Latent Reasoning via Hidden State Recycling - Publications - ByteDance Seed Team

Recycler introduces a lightweight adapter that passes both embeddings and the previous token's hidden state to the next step, preserving cross-step semantics with minimal overhead. It achieves strong parameter efficiency and competence in reasoning tasks, performing well even when added later in training and outperforming larger models.

Artificial intelligence learned to rummage through the Finnish waste mound, but it still cannot remove the human role in sorting | Domestic news | Yle

Translated title. Original: Tekoäly oppi penkomaan suomalaista jätevuorta, mutta ihmisen roolia lajittelussa se ei pysty vielä poistamaan | Kotimaa | Yle

AI is trained to identify Finnish waste to improve sorting for a waste-to-energy plant. Despite high-speed image analysis and a country-specific dataset, human input is still essential for effective pre-sorting and final decisions; the project aims to support EU circular economy goals and data sharing for further research.

Rare Earths and Lithium: How Biotechnology Is Changing Metal Recycling | heise online

Translated title. Original: Seltene Erden und Lithium: Wie Biotechnologie das Metallrecycling verändert | heise online

The article describes research using biomolecules and bacteria to recover rare earths and lithium from wastewater and other sources. It also reports on a Pforzheim plant testing a way to recycle high-performance magnets while retaining their valuable alloy.

Artificial intelligence is already in the sorting facility. Two picks per second and 90% accuracy.

Translated title. Original: Sztuczna inteligencja już w sortowni. Dwa chwyty na sekundę i 90 proc. skuteczności

A waste-sorting plant near Třinec is testing an AI system that combines mechanical sorting with two robots, which can make up to two picks per second. The facility is expected to process 20,000 tonnes of waste a year and begin full operation on January 1, 2027.