Theia — Automated Crate Inspection System
Fresh produce and poultry supply chains, Australia
A neural-network-powered vision system inspecting nine sides of each crate simultaneously, deployed across fresh produce and poultry supply chains at scale.
- 5x Industrial cameras
- Neural network classification
- Custom HMI
- Real-time defect detection
- Food-grade environment rated
The Situation
Food distribution operators rely on reusable plastic crates that cycle through wash processes before re-entering the supply chain. The problem was inspection. Manual assessment was inconsistent, slow, and produced two failure modes that were both costly: missed contamination reaching the supply chain, and excessive false rejections sending clean crates back for unnecessary rewashing.
A reliable, objective inspection method was needed that could operate at line speed without adding manual labour.
The Work
React Automation developed Theia, a purpose-built crate inspection system using five industrial cameras to simultaneously assess nine sides of each crate. A trained neural network processes the imagery in real time, distinguishing genuine contamination from acceptable surface artefacts that a rules-based system would have incorrectly flagged.
The result is a system that makes consistent, objective pass/fail decisions at production speed, with an HMI providing operators full visibility of defect detection across each crate face.
The Result
Missed contamination fell by over 95%. False rejection rates dropped substantially, reducing unnecessary rewash cycles and the operational costs that came with them. Food safety compliance outcomes improved, and the system provided operators with a level of auditability that manual inspection could never deliver.
Theia has now inspected more than 8 million crates in live production.
Want to know if you can improve your efficiency?
We'd like to hear about your supply chain and see how we can save you time and increase your bottom line.