Date published: December 16, 2025​​

John Ilett, President and Founder of Emergent Vision Technologies describes the company’s next steps in high-performance machine vision, focusing on how cameras, software, and GPU-accelerated processing work together to enable scalable, real-time systems. The talk combines product overviews with live demonstrations to show how complex vision applications can be built and deployed efficiently.

Camera and Hardware Focus

John highlights Emergent’s expanding camera portfolio.

• Introduction of the EROS 10GigE camera line, featuring a compact 29 × 29 mm form factor and very low power consumption

• BOLT 25GigE and ZENITH 100GigE cameras, including high-resolution area-scan and ultra-fast line-scan and TDI models

• Support for SWIR sensors, modular transceivers, and multiple lens mount options

Software and System Architecture

The presentation explains how Emergent’s software removes traditional data-transfer bottlenecks.

• Zero-copy and GPU Direct image transfer on Windows and Linux

• FlexProc and FlexTrans frameworks for flexible processing and data movement

• Use of eSDK, eSDK Pro, and eCapture Pro to simplify development and scaling

Demonstrations and Use Cases

Several live demos show the system in action.

• Multi-camera setups with eight cameras per server and distributed GPU processing

• Real-time GPU tasks such as image processing and H.265 compression

• A full PCB inspection workflow using EROS cameras, rotary stages, LED projectors, and unsupervised AI anomaly detection to identify and highlight defects