AVerMedia debuts robotics platform lineup for drones, AMRs and humanoids

AVerMedia Technologies unveiled a new robotics platform portfolio on September 16, 2026, in New Taipei City to help developers build physical AI systems for drones, autonomous mobile robots and humanoids. The lineup is built around NVIDIA Jetson modules and adds camera, I/O, BSP and memory-optimization support aimed at reducing integration work.

Why it matters: - AVerMedia’s new robotics platform portfolio is aimed at lowering the integration burden as physical AI systems add more sensors, cameras, control paths and AI workloads. - The company is targeting three fast-growing robotics categories: drones, autonomous mobile robots and humanoids. - The platforms are designed to help developers move faster from development to deployment with less custom board work.

What happened: - AVerMedia Technologies announced the portfolio on September 16, 2026, in New Taipei City, Taiwan. - The lineup includes three application-focused platforms: D137 Lynx for drones, D317 Jaguar for AMRs and D333 Panther for humanoid robotics. - The portfolio sits under AVerMedia’s “Lead the Next Wave of Physical AI” initiative.

The details: - D137 Lynx targets drones and UAVs and uses NVIDIA Jetson Orin Nano and Orin NX in a 90 × 60 mm carrier that weighs 50 g. - D137 Lynx adds flexible camera connectivity and integrated control I/O to cut down on extra boards and wiring. - D137 Lynx supports MIPI, gimbal and GMSL camera architectures. - D317 Jaguar combines NVIDIA Jetson AGX Orin with high-bandwidth sensor expansion, localization and control connectivity, storage, wireless communications and optional remote-management capabilities. - D317 Jaguar includes flexible I/O expansion, including GMSL and PoE connectivity options. - D333 Panther brings NVIDIA Jetson Thor T4000 and T5000 to a 120 × 120 mm carrier for sensor-rich humanoid systems. - D333 Panther includes a modular design with the F8G7C daughter board for cameras, control systems, USB devices and networking. - AVerMedia supports integration with RealSense and Stereolabs depth cameras across the robotics portfolio. - AVerMedia provides both Standard and Real-Time BSPs, released at the same time. - The Real-Time BSP includes PREEMPT_RT support for more predictable, low-latency execution. - AVerMedia supports NVIDIA memory-optimization technologies for Jetson Thor, Jetson AGX Orin and Jetson Orin Nano/NX. - The company said those optimizations can reduce memory footprint and may let customers right-size their platform within the Jetson family while keeping required performance.

Between the lines: - The portfolio is built around the idea that robotics developers need more than compute chips; they need application-specific hardware, software and peripheral integration that work together. - The emphasis on BSP customization, validation and production support suggests AVerMedia wants to compete as a full-stack enablement partner, not just a board supplier. - The product split tracks three different robotics form factors with different constraints: airframe weight for drones, sensor growth for AMRs and compact high-density compute for humanoids.

What's next: - AVerMedia said it will support developers with hardware, BSP customization, peripheral integration, validation and production support as projects move toward deployment. - The company is positioning the portfolio for future robotics programs that need scalable sensing, timing and memory efficiency. - Alex Liu, vice president of AVerMedia, said the portfolio is intended to help developers integrate faster, scale more easily and move confidently toward deployment.

The bottom line: - AVerMedia is betting that the next stage of physical AI will reward robotics platforms that reduce integration friction while keeping compute, sensing and timing flexible enough for real-world deployment.