Advantech, a global leader in edge computing, is bringing unprecedented AI performance to the edge with the introduction of a new lineup of application-focused solutions powered by the NVIDIA Jetson Thor modules. These new modules set a new benchmark for edge AI, delivering up to 2070 FP4 TFLOPS of performance. Advantech's solutions translate this raw power into practical, hardware-software integrated systems specifically designed for robotics, medical AI, and data AI applications, aiming to accelerate development and deployment for engineers and enterprises.
Advantech launches new Edge AI solutions powered by the high-performance NVIDIA Jetson Thor modules.
The solutions are tailored for three key areas: robotics, medical AI, and data AI.
Each offering combines application-specific hardware with pre-integrated software stacks like Robotic Suite and GenAI Studio.
The platforms are built on a container-based architecture for flexibility and faster development.
Key features include support for multi-camera GMSL, real-time sensor processing, and low-latency data flows.
The Advantech Container Catalog provides ready-to-deploy edge AI applications.
For the robotics sector, Advantech introduces the ASR-A702 and AFE-A702 controllers, designed for humanoid robots, Autonomous Mobile Robots (AMRs), and unmanned vehicles. These systems deliver real-time AI reasoning and GPU-accelerated Simultaneous Localization and Mapping (SLAM). They support multiple GMSL cameras and 2D/3D sensors, and are pre-integrated with the Robotic Suite for plug-and-play development. With features like hardware time synchronization, anti-vibration design, and over-the-air (OTA) upgrades, they are built for rapid integration and reliable operation in dynamic environments.
In the medical field, Advantech is empowering next-generation applications with its AIMB-294 board and EPC-T5294 system. Leveraging NVIDIA Jetson Thor alongside advanced SDKs like Holoscan and MONAI, these platforms are engineered to accelerate real-time sensor processing, image analysis, and AI-driven diagnostic pipelines. They are optimized for pre-trained models and 3D imaging, providing the low latency and high precision required for surgical robotics, clinical workflows, and intelligent diagnostic tools.
For data-intensive AI applications, the AIR-075 system is designed to handle demanding tasks like Visual Language Model (VLM) inference and multi-camera AI vision analysis. It features four 10GbE and GMSL interfaces to manage high data throughput in traffic and factory settings. When combined with the NVIDIA Metropolis and Triton inference server, along with Advantech's own Edge AI SDK, it enables sophisticated sensor fusion, multi-model inference, and centralized management for predictive, real-time edge intelligence.
A key enabler across all solutions is the Advantech Container Catalog (ACC). This resource provides a cluster of ready-to-develop edge AI applications, including end-to-end computer vision and Edge LLM environments that are optimized for AI agent integration. Fully compatible with Advantech's WISE-Edge Developer Architecture (WEDA), the containerized approach allows for scalable expansion from single-node setups to distributed edge networks, significantly speeding up the development cycle.
TechIntelPro empowers B2B technology decision-makers with the latest industry trends, deep insights, leadership perspectives and actionable intelligence to drive both organizational and professional milestones—and thrive on the tech pulse.