Robots Could Turn E-Waste Into a Source of Legacy Chips

2026-05-19 · Source: IEEE Spectrum · Field: Technology & Digital — Robotics & Autonomous Systems, Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Intermediate, short

Summary

Tuurny, a San Francisco-based startup, is developing an automated robotic system called Nantul to recover reusable chips from electronic waste, addressing increasing regulatory pressure and the projected 82 million tonnes of e-waste annually by 2030. The United Nations' 2024 Global E-Waste Monitor reported that current management captures less than a third of recoverable metal value. Nantul is designed to identify and extract RAM integrated circuits, with each machine capable of recovering 300 intact RAM ICs per hour. Tuurny secured a six-figure deal with UK television recycler Areera, processing 1,500 tonnes per month, for a field deployment planned for early 2027. The system targets legacy chips, which are difficult to source, and aims to create a new supply chain by separating components before shredding, contrasting with conventional recycling that often destroys reusable parts. The four-person startup also received a NASA-funded grant for an AI-powered repair assistant.

Key takeaway

For AI Engineers or Directors of AI/ML evaluating e-waste recycling solutions, Tuurny's robotic disassembly approach offers a promising strategy to recover valuable legacy chips and critical materials. You should consider how automated, component-specific recovery systems could integrate into your supply chain to mitigate sourcing challenges for older hardware. This method reduces waste and creates new material streams, potentially addressing both environmental compliance and critical component availability.

Key insights

Automated robotic disassembly of e-waste can recover valuable legacy chips and materials before bulk shredding.

Principles

• Prioritize component recovery before bulk processing.
• Target high-value, standardized components first.
• Adaptability to e-waste variability is crucial.

Method

Nantul uses a neural network, computer vision, controlled heat, and robotic controls to identify, remove, and sort chips based on manufacturers' thermal profiles, minimizing damage.

In practice

• Focus on RAM ICs for high reuse value.
• Supply recovered aluminum and copper to smelters.
• Build modular machines with off-the-shelf parts.

Topics

• E-waste Recycling
• Robotic Disassembly
• Legacy Chips
• Supply Chain Resilience
• Computer Vision
• RAM Recovery

Best for: Computer Vision Engineer, Investor, CTO, AI Engineer, Robotics Engineer, Director of AI/ML

Related on AIssential

• AI data center boom is leaving consumer electronics short of chips − even though they don’t use the same kinds — AI data center demand is reorganizing the chip market, creating shortages and higher costs for consumer electronics.
• Rethinking Digital Infrastructure: The Rise of Zero Data Waste In an era dominated by cloud… — Digital infrastructure can pivot from data destruction to on-site processing and energy generation, reducing waste and costs.
• This startup wants to dismantle EVs to solve Europe’s critical materials problem — Automated EV dismantling using AI and robotics can secure critical material supply chains and meet recycling targets.
• With aluminum prices up 20%, recycling startups bet on AI to cash in — AI-powered sorting significantly boosts aluminum recycling efficiency and domestic supply amidst rising prices and critical mineral status.
• LabCycle secures £430K to commercialise lab plastic recycling system and cut incineration waste — LabCycle's AutoDecon system recycles contaminated lab plastics into high-grade materials, fostering a circular economy and cutting emissions.
• Meet the Speakers of The Future of AI Panel @LCAW — AI is transforming sustainability by enabling advanced solutions for environmental management and circular economy initiatives.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by IEEE Spectrum.