Industry Pulse News Desk · Edited by Viktor Peshekhonov · · 6h ago
Redwood Materials Deploys AI-Powered Battery Sorting System
An automated vision and sensor system identifies and routes tons of mixed batteries per hour with 98 percent accuracy.
Redwood Materials has deployed an artificial intelligence-driven sorting system designed to categorize spent batteries by chemistry and form factor at scale, according to the recycling firm.
The automated processing line handles several tons of mixed battery inventory per hour. Inbound batteries delivered in bulk containers are mechanically separated into a single-file feed before passing through a specialized inspection tunnel equipped with high-resolution cameras, optical scanners, and weight sensors.
An internally developed machine-learning model, trained on proprietary battery datasets and running on an on-site graphics processing unit cluster, evaluates each unit's label, dimensions, and weight. The system then routes the battery to its designated processing stream, achieving an identification accuracy rate above 98 percent.
Initial sorting represents a persistent operational bottleneck for commercial recyclers due to the proliferation of diverse consumer and industrial battery formats. Commingling incompatible chemical compositions lowers recovery efficiency, elevates operational hazards, and drives up downstream processing expenses.
The deployment supports Redwood Materials' broader initiative to expand domestic recovery of high-value battery metals, including lithium, nickel, and cobalt. The company, established by former Tesla co-founder JB Straubel, processes manufacturing scrap and end-of-life packs to supply refined materials back into the electric vehicle supply chain.