🍎Apple Machine Learning•Feb 25, 2026Stalecollected in 19h

Apple's A.R.I.S. AI Sorts E-Waste Real-Time

Post LinkedIn

🍎Read original on Apple Machine Learning

#computer-vision#e-waste#object-detection#sustainabilitya.r.i.s.appleyolox

💡Apple's 90% accurate CV sorter for e-waste—ideal blueprint for industrial vision apps.

⚡ 30-Second TL;DR

What Changed

Employs YOLOx for real-time e-waste classification

Why It Matters

Boosts e-waste recycling efficiency, reducing resource loss. Demonstrates practical CV applications in sustainability for AI practitioners.

What To Do Next

Integrate YOLOx into your CV pipeline for real-time object sorting prototypes.

Who should care:Researchers & Academics

Key Points

•Employs YOLOx for real-time e-waste classification
•Classifies metals, plastics, circuit boards
•90% overall precision, 82.2% mAP, 84% sortation
•Low-cost, portable design for shredded e-waste
•Addresses traditional recycling inefficiencies

🧠 Deep Insight

Background and context from public sources — not the original article. 8 sources cited.

🔑 Enhanced Key Takeaways

•A.R.I.S. achieves stable real-time performance exceeding 20 FPS on Mac mini using CoreML acceleration.
•The system integrates iterative data augmentation and model-in-the-loop refinement for improved detection of plastics and small fragments.
•A.R.I.S. complements Apple's prior recycling robots like Daisy, which disassembles iPhones, and sorting machines Dave and Taz deployed in China.

🛠️ Technical Deep Dive

•Inference pipeline includes steps optimized for low-cost hardware, avoiding unnecessary scaling to preserve resolution.
•Utilizes YOLOx model with CoreML acceleration on Mac mini, delivering over 20 FPS for real-time shredded e-waste sorting.
•Employs iterative data augmentation and model-in-the-loop refinement processes to enhance accuracy on challenging materials like plastics and small fragments.

🔮 Future ImplicationsAI analysis grounded in cited sources

A.R.I.S. will be shared with global recycling partners as a low-cost solution.

Apple commits to sharing the technology to lower barriers to advanced recycling adoption and create industry-wide impact.

System improvements will target higher accuracy for plastics and small fragments.

Ongoing refinements using data augmentation and model-in-the-loop processes address current limitations in these categories.

⏳ Timeline

2016-03

Introduced Liam robot for iPhone disassembly, processing 1.2 million units per year.

2018-04

Debuted Daisy robot, disassembling 200 iPhones per hour across multiple models.

2022-01

Expanded Daisy to handle 18 iPhone models with 15 material output streams.

2024-01

Deployed Dave and Taz recycling machines with partner in China; introduced new product sorter in California.

2025-01

Daisy updated to disassemble 29 iPhone models.

2026-02

Published A.R.I.S. paper introducing AI-powered sorter for shredded e-waste.

📎 Sources (8)

Factual claims are grounded in the sources below. Forward-looking analysis is AI-generated interpretation.

🍎Read original article on Apple Machine Learning

📰

Weekly AI Recap

Read this week's curated digest of top AI events →

👉Related Updates

Same topic

Explore #computer-vision

Same product

DeepSeek Finally Unveils Vision Model

cnBeta (Full RSS)•Aug 22

AI-curated news aggregator. All content rights belong to original publishers.
Original source: Apple Machine Learning ↗

This is a summary, not the original. Read the source, or get the weekly briefing.

Weekly AI briefing

One email a week. Unsubscribe anytime.