Edge AI provider Aetina will showcase new platforms combining 3D vision and high-performance edge computing at NVIDIA GTC 2026, targeting emerging physical AI applications in robotics and industrial automation.
At the event, Aetina will demonstrate its DeviceEdge AIB-AT78/68 series, powered by NVIDIA Jetson Thor modules built on **NVIDIA’s Blackwell architecture. The platform is designed for robotics and industrial automation applications that require real-time perception, high-bandwidth sensor integration, and edge-based AI processing.
As part of the demonstration, Aetina is collaborating with Stereolabs to present a humanoid surround-vision system integrating stereo and monocular cameras. The setup combines ZED X stereo cameras for frontal depth sensing with ZED X One cameras for peripheral awareness, enabling robots to generate real-time depth maps and environmental awareness for navigation and manipulation tasks.
The platform also supports high-speed connectivity and sensor fusion capabilities aimed at enabling more complex robotic behaviours in sectors such as manufacturing, logistics, and service robotics.
Alongside the robotics-focused solution, Aetina will also present its SuperEdge AEX-2UA1 edge server, based on the NVIDIA MGX architecture and accelerated by RTX PRO 6000 Blackwell GPUs. The system is designed to support enterprise AI agents and large language model inference at the edge, enabling organisations to analyse operational data and generate real-time insights locally while maintaining data sovereignty.
Edge platforms capable of combining perception, sensor fusion, and AI inference locally are becoming increasingly important as robotics systems move from controlled environments toward more dynamic industrial deployments.
Aetina’s AIB-AT78/68 platform is expected to become commercially available in March 2026.
This article was adapted from a company press release and edited for publication.
To learn more, visit the Aetina website
