www.ptreview.co.uk
17
'26
Written on Modified on
Integrating Edge AI Architecture and Stereo Vision for Industrial Automation
SINTRONES Technology and Stereolabs have integrated rugged computing platforms with global shutter cameras to enhance real-time depth perception in autonomous systems.
www.sintrones.com
Industrial automation, autonomous vehicles, and defense applications require continuous real-time spatial awareness when operating in harsh, dynamic environments. Standard computing systems and optical sensors frequently fail under extreme temperatures, mechanical vibrations, and moisture exposure. To address these operational challenges, SINTRONES Technology Corp. and Stereolabs have pooled their respective expertise in hardware ruggedization and advanced spatial perception to deliver a unified edge AI vision system.
Technical Solution and Functional Architecture
The architecture combines the SINTRONES IBOX-650P-IP66-2S/DIO edge computer with Stereolabs ZED X and ZED X One cameras. Central processing is driven by the NVIDIA Jetson AGX Orin module, which delivers up to 275 TOPS of artificial intelligence performance dedicated to localized inference and navigational data processing.
Hardware responsibilities are split between the two providers to optimize system compatibility:
- Sintrones Computing Infrastructure: The IBOX platform provides a fanless, IP66-rated enclosure that protects the internal electronics from dust and high-pressure water jets, ensuring thermal dissipation without external air exchange.
- Stereolabs Sensor Integration: Spatial data collection relies on the ZED X stereo camera for high-precision depth sensing and 3D perception. It features built-in self-calibration to maintain measurement accuracy during physical shocks. For high-speed motion tracking, the ZED X One global shutter camera captures unblurred images in vibration-prone environments.
Data transmission between the cameras and the edge computer utilizes Gigabit Multimedia Serial Link (GMSL) interfaces. This connection standard ensures low-latency, high-bandwidth video transmission over extended physical distances, which is critical for mobile machine vision deployments. Hardware synchronization across the GMSL interfaces aligns multi-camera frame captures precisely to prevent telemetry errors.
Deployment and Target Applications
The combined digital infrastructure is designed for immediate integration into autonomous mobile systems, heavy robotics, and intelligent surveillance setups. The system installs directly onto mobile platforms, interfacing with existing vehicle control networks via dedicated digital input/output (DIO) lines.
The primary deployment use case involves vehicle navigation and threat localization, where the Stereolabs cameras feed raw spatial data directly into the Jetson platform. The local AI model processes this data instantly to calculate obstacle distances and coordinate paths, eliminating the latency and security risks of cloud-based processing. This localized architecture improves process stability and maintainability by minimizing external network dependencies.
The integrated vision solution was demonstrated at Eurosatory 2026 in Paris, France, from June 15 to June 19, 2026, showcasing functional real-time 3D mapping and edge inference capabilities.
Edited by Evgeny Churilov, Induportals Media - Adapted by AI.
www.sintrones.com
Deployment and Target Applications
The combined digital infrastructure is designed for immediate integration into autonomous mobile systems, heavy robotics, and intelligent surveillance setups. The system installs directly onto mobile platforms, interfacing with existing vehicle control networks via dedicated digital input/output (DIO) lines.
The primary deployment use case involves vehicle navigation and threat localization, where the Stereolabs cameras feed raw spatial data directly into the Jetson platform. The local AI model processes this data instantly to calculate obstacle distances and coordinate paths, eliminating the latency and security risks of cloud-based processing. This localized architecture improves process stability and maintainability by minimizing external network dependencies.
The integrated vision solution was demonstrated at Eurosatory 2026 in Paris, France, from June 15 to June 19, 2026, showcasing functional real-time 3D mapping and edge inference capabilities.
Edited by Evgeny Churilov, Induportals Media - Adapted by AI.
www.sintrones.com
