UnitX launched DeteX on June 15, an AI-powered smart camera the company claims is the fastest to deploy in its category, with a five-step guided setup that puts operators live in one minute. The Milpitas, California company inspects more than $15 billion in products annually across more than 190 factories worldwide and says DeteX brings enterprise-grade deep learning to production lines at the price of a basic vision sensor.
The product targets what UnitX calls the "Vision Gap," the divide between low-cost sensors that fail under real-world variability and enterprise vision systems that require specialized integrators, proprietary software certifications, and months of commissioning work. Where competing AI vision systems take weeks to integrate, DeteX requires no vision engineering background to operate.
What's new here is the feature depth packed into a single edge device: pixel-precise AI segmentation, multi-class defect classification, OCR that handles engraved, low-contrast, curved, and dirty characters at line speed, and a counting tool that tracks up to 999 targets per cycle within user-defined regions. Dimensional measurement reaches 0.1mm accuracy on diameter, edge, line, and pitch, a threshold UnitX says standard 1.6-megapixel sensors cannot achieve. The camera is ecosystem-agnostic and connects to existing production lines without proprietary hardware requirements.
"Basic sensors fail when conditions vary, and enterprise systems demand months of complex integration," said Keven Wang, UnitX CEO. "With DeteX, we've built an edge-AI camera that gives engineers the best of both worlds."
DeteX became commercially available June 15 and made its first public showing at Automate 2026 in Chicago, June 22-25, at McCormick Place, Booth 13000. UnitX has not disclosed a unit price. Sectors targeted include automotive, medical device assembly, food and beverage packaging, and electronics and PCB inspection. The company specifically highlights PCB applications, noting the camera detects partially seated connectors and micro-surface anomalies that conventional presence/absence sensors cannot resolve.
