Advantech Introduces UniCAM to Unify Multi-Camera Integration for Robotics and Physical AI

Industry’s first cross-brand GMSL2 camera integration architecture enables flexible camera mix-and-match, intelligent configuration, hardware synchronization, and ROS 2 integration to reduce POC cycles from weeks to days

Taipei, Taiwan, November 6, 2026 – Advantech, a global leader in edge computing and industrial AI, today introduced Advantech UniCAM, a unified camera integration architecture built around GMSL2 (Gigabit Multimedia Serial Link 2) technology from Analog Devices, Inc. (ADI), designed to simplify multi-camera development for robotics, humanoids, autonomous machines, and Physical AI applications.

“As robotic vision systems become increasingly complex, developers need greater flexibility to integrate different cameras and computing platforms. Advantech UniCAM moves repetitive low-level integration into a unified architecture, reducing camera compatibility and configuration effort so developers can focus more on perception and Physical AI applications and accelerate the path from development to deployment,” — Nomo Hsu, Robotics Software Director, Advantech.

“Advantech’s unified integration architecture demonstrates the value of ecosystem collaboration in simplifying GMSL-based multi-camera system development. By enabling cameras from multiple vendors to operate through a common platform, Advantech UniCAM helps reduce integration complexity and gives developers greater flexibility in building robotics, machine vision, and other intelligent systems. Together, we are helping broaden access to GMSL technology and accelerate its adoption across these emerging applications,” said Prem Nayar, Head of Automotive Video and Data Solutions at ADI.

“Machine vision is a critical foundation for intelligent robots and Physical AI systems. By streamlining multi-camera configuration, synchronization, and ROS 2 integration, Advantech UniCAM can help developers move more efficiently from hardware integration to perception and application development. This is an important step toward accelerating scalable robotics innovation and real-world deployment,” said Philippe Reiber, Global Head of Industrial Sales at ADI.

Addressing a Critical Challenge in Robotic Vision

As robots and humanoids become increasingly vision-centric, real-time perception depends on fast image transmission and precise synchronization across multiple cameras. GMSL has become an important interface for robotic vision, providing the bandwidth and deterministic connectivity needed for demanding multi-camera applications.

However, GMSL drivers have traditionally been developed independently by individual camera vendors, limiting the flexibility to combine different brands within the same SerDes architecture. Robot developers therefore often prioritize GMSL for synchronization-sensitive devices such as 3D and stereo cameras, while using Ethernet or USB for additional cameras, resulting in a more fragmented system design.

Working with leading camera ecosystem partners, Advantech developed UniCAM to enable supported GMSL cameras from different vendors to operate within the same SerDes architecture. Developers can now mix and match cameras through a common interface while simplifying hardware synchronization, configuration, and system integration.

As the industry’s first cross-brand GMSL2 camera integration solution built on a unified architecture, UniCAM reduces repetitive driver adaptation and low-level configuration, helping shorten POC cycles from weeks to days and accelerate robotics application development.

One Scalable Architecture, Expanding Camera Ecosystem

UniCAM replaces vendor-specific integration with a scalable architecture that centrally manages supported cameras and shared hardware resources while preserving each vendor’s original functionality, SDKs, algorithms, and image quality.

The architecture currently integrates solutions from Stereolabs, RealSense, Orbbec, Sensing World, and Leopard Imaging, supporting flexible combinations of stereo, depth, and 2D cameras.

Advantech will continue expanding the UniCAM ecosystem with additional camera partners and models, allowing new cameras to be added without rebuilding the underlying integration stack.

Intelligent Camera Detection and Configuration

UniCAM further simplifies deployment through intelligent camera detection and automatic configuration within the Advantech Robotic Suite development environment.

When a supported camera is connected, UniCAM automatically identifies the camera and generates the corresponding device settings, reducing manual device-tree modification and low-level setup.

Through the GMSL/MIPI Camera Selector and Camera Configuration Dashboard, developers can manage supported cameras in a common environment and move efficiently from initial setup to ROS 2 development.

Accelerating ROS 2 Application Development

Pre-integrated camera wrappers and configurable background services enable supported cameras to connect directly to ROS 2 environments, with camera topics automatically initialized at system startup.

Together with Advantech Robotic Suite, developers can move from camera integration toward perception, SLAM, navigation, manipulation, and Physical AI applications through a more streamlined hardware-to-application workflow.

Scaling Across Cameras and Computing Platforms

The architecture initially supports NVIDIA Jetson platforms, including Jetson AGX Thor, with support extending to AGX Orin and Orin NX. The same approach is planned for Intel IPU7 and Qualcomm Dragonwing IQ9 platforms.

Advantech is building toward an N Cameras × N Computing Platforms model, allowing developers to select cameras based on imaging and sensing requirements while independently choosing the computing platform that best fits their performance and power needs.

As support expands across both sides of the ecosystem, UniCAM reduces repeated integration effort and helps preserve software investments across robotics platforms and product generations.

From Camera Integration to Application Innovation

Advantech brings together robotics computing, GMSL2 connectivity, intelligent configuration, hardware synchronization, ROS 2 integration, and an expanding camera ecosystem into a scalable robotic vision foundation.

By moving repetitive low-level integration into a reusable platform layer, developers can spend less time resolving camera compatibility and more time building differentiated perception and Physical AI applications.

UniCAM provides a scalable path from camera integration to application innovation, helping accelerate the next generation of intelligent robots and Physical AI systems.

Explore Advantech UniCAM and Robotic Suite.

About Advantech

Advantech is a global leader in intelligent IoT systems and embedded platforms, with the corporate vision of “Enabling an Intelligent Planet.” To embrace the trends of edge computing and artificial intelligence, Advantech has adopted a sector-driven strategy centered on Edge Computing and Edge AI. The company focuses on five key markets: edge intelligent systems, smart manufacturing, energy and utilities, intelligent healthcare, and intelligent retail and services. By integrating its Edge Computing hardware platforms, WISE-IoT industrial IoT software platform, industry-specific Edge AI solutions, and domain expertise, Advantech enables collaborative business models for integrated industry applications and helps partners connect across the value chain. Advantech also actively co-creates industry ecosystems with partners to accelerate industrial intelligence. (www.advantech.com)