UnitX Launches DeteX: An AI-Powered Smart Camera for Simplified Manufacturing Inspection

DeteX leverages a compact AI model for robust defect detection, classification, and measurement, with minimal retraining needed, and supports standard industrial communication protocols for seamless integration.

Key Highlights

  • DeteX features an 8 MPixel sensor with 30 fps frame rate, IP54 rating, and supports major PLC protocols for easy integration into existing factory setups.
  • The AI segmentation model learns at a pixel level, tolerating natural variations and requiring as few as five new images for retraining in significant condition changes.
  • Designed for simple deployment, DeteX allows line technicians to set up and configure the system in under a minute without specialized training.
  • While optimized for in-process inspection, DeteX is not suitable for applications needing multi-camera coverage or high-precision measurements on reflective surfaces.
  • The camera expands opportunities for manufacturers and integrators by providing powerful AI inspection capabilities in a user-friendly package, broadening the automation market.

UnitX (Milpitas, CA, USA), a robotics company specializing in AI-powered visual inspection in manufacturing, has released an AI-powered smart camera, DeteX. The company debuted the camera at Automate 2026, held June 22-25 in Chicago.

According to UnitX, the 8 MPixel, 30 fps frame rate, IP54 rated camera combines AI-powered inspection, dimensional measurement, OCR, classification, and counting in a single device that the company says can be deployed in one minute. The camera does not require vision engineering experience; it is designed to be easily set up and deployed. It is especially designed for inspection applications in industries such as automotive manufacturing, medical device assembly, food and beverage packaging, industrial assembly, and electronics and PCB assembly.

Vision Systems Design wanted to learn more about DeteX, so we reached out to Kevin Wang, co-founder and CEO of UnitX. 

Editor’s note: The following Q&A may have been edited for style and/or clarity.

Vision Systems Design (VSD): How does DeteX's AI segmentation perform in highly variable production environments, and what retraining is required when conditions change?

Kevin Wang (KW): DeteX's AI segmentation is a small AI model (as distinct from Large Language Models) and learns the visual signature of acceptable and defective parts at a pixel level, giving it inherent tolerance for natural variation such as defect variations, inconsistent lighting, surface texture differences, and part-to-part variability that would break rules-based systems.
When conditions change significantly, the model can be updated with as few as five new sample images. For minor variation within the same defect type and product family, no retraining is typically required.

For high-mix production environments, UnitX's feature-centric AI architecture takes this further. Rather than training separate models for each product, UnitX AI uses a shared defect feature library that trains on the same defect type across multiple materials and product families. When a defect feature is updated or improved, the change cascades automatically to all models that reference it—faster changeovers, consistent results, and less duplicate training work.

VSD: How open is the DeteX ecosystem, and what integration standards does it support?

DeteX supports major PLC protocols including EtherNet/IP, PROFINET, as well as TCP/IP for MES, FTP, and custom applications, and direct I/O output for triggering line hardware without a PLC in the loop. PLC recipe switching allows different inspection configurations to be activated directly from the line controller.

It’s designed to fit existing factory infrastructure without disruption, using the same communication standards, network topology, and PLC logic already in use.

VSD: What applications are DeteX best suited for? When would a full multi-camera system be the better choice?

DeteX is purpose-built for in-process inspection: presence/absence checks, classification, surface defect detection, OCR, barcode reading, dimensional measurement, and counting in a situation where a basic sensor is insufficient, but a full enterprise system is an overkill.

A multi-camera or enterprise system is the right choice when applications require simultaneous inspection from multiple angles and detection on complex geometry, fly capture at high line speeds, or custom deep learning architecture.

VSD: Are there any sacrifices made in capabilities versus the simplicity of setup?

KW: The AI capability is not simplified. The AI segmentation, classification, OCR, and measurement tools are built on the same underlying technology as our enterprise platform and they are packaged for deployment by a line technician in under a minute, without specialist involvement.

What DeteX is not designed for is applications requiring multi-camera coverage or micrometer detection on highly reflective geometry without supplementary lighting. It covers everything else for in-process applications.

VSD: Do you believe AI smart cameras will replace traditional machine vision integration projects, or create new opportunities for integrators?

KW: DeteX unlocks vision applications which manufacturers could not find a solution between basic vision sensors vs. complex vision systems. It brings powerful vision capability at the ease of use of a sensor.

The larger opportunity is expansion. Integrators who have historically avoided vision projects can now take them on, with DeteX handling the vision component, expanding the total addressable market for SI-delivered automation. Complex applications such as multi-camera systems, high-speed inspection, micron-level precision still require deep expertise and full enterprise solutions. DeteX solves the in-process use cases with an easy-to-use user interface that can bring more customers into the vision inspection market.

About the Author

Jim Tatum

Senior Editor

VSD Senior Editor Jim Tatum has more than 25 years experience in print and digital journalism, covering business/industry/economic development issues, regional and local government/regulatory issues, and more. In 2019, he transitioned from newspapers to business media full time, joining VSD in 2023.

Sign up for our eNewsletters
Get the latest news and updates

Voice Your Opinion!

To join the conversation, and become an exclusive member of Vision Systems Design, create an account today!