Company News

Highlights from the NVIDIA Jetson Embodied‑Brain Mass‑Deployment Practice Seminar

On September 17, 2026, the “NVIDIA Jetson Embodied‑Brain Mass‑Deployment Practice Seminar,” hosted by RealTimes Technology (Beijing) Co., Ltd. (RealTimes), was successfully held at the CIMC Entrepreneur Exchange Center in Nanshan District, Shenzhen. Focused on taking embodied intelligence from algorithm prototyping to scalable mass production, the event brought together NVIDIA technical experts, technical leaders from leading industry companies, and professionals across the robotics value chain. Discussions covered frontier topics such as physical AI, the Jetson Thor platform, engineering deployment of embodied “brains,” and new perception solutions. An on‑site exhibition zone showcased real Jetson‑based embodied‑brain hardware, combined with networking breaks, so that theoretical sharing and hands‑on experience went hand in hand to explore the path to commercializing embodied intelligence.

Experts Convene to Discuss Breakthroughs in Embodied‑Intelligence Mass Production

The seminar opened with a welcome address by Mr. Wang Guangming, General Manager of RealTimes Technology (Beijing) Co., Ltd. Mr. Wang welcomed all guests, shared insights into the current development opportunities in embodied intelligence, and outlined RealTimes’ commitment to the NVIDIA Jetson ecosystem—focusing on embodied‑brain hardware solutions and helping robotics companies overcome mass‑deployment challenges—setting the stage for the technical program.

Featured Speaker Sessions

NVIDIA Empowering Physical AI: Generative AI Reshaping the Embodied‑Intelligence Industry — Du Juan, Solutions Architect, NVIDIA Robotics

Starting from the fundamentals of physical AI, Du Juan introduced NVIDIA’s robotics technology stack, explained how generative AI is reshaping perception, planning, and control pipelines for embodied intelligence, analyzed the technical challenges of humanoid and quadruped robots operating in real physical environments, and demonstrated how NVIDIA’s full software‑hardware stack enables robots to learn and make decisions in the physical world, giving engineers a forward‑looking technical perspective.

NVIDIA Jetson Thor and HSB Accelerating Real‑Time Inference for Embodied Intelligence — Lei Hao, Developer Relations Manager, NVIDIA

Lei Hao highlighted the new‑generation Jetson Thor platform and HSB hardware solution, comparing improvements in compute, power efficiency, and I/O. For real‑time embodied‑robotics inference, he explained how Thor’s powerful compute supports multimodal perception and on‑device large‑model inference, addressing pain points such as high edge‑inference latency and limited resources, and demonstrated the platform’s engineering advantages in embodied‑intelligence tasks.

Jetson Embodied‑Brain Mass‑Deployment Practice — Gao Feng, Product Manager, RealTimes

From a hardware and mass‑production perspective, Gao Feng detailed end‑to‑end embodied‑brain solutions based on the Jetson series. Covering carrier‑board design, multi‑sensor interfacing, thermal and reliability design, batch manufacturing, factory testing, and long‑term operations, he broke down common engineering pitfalls in production projects and shared RealTimes’ delivery experience from numerous robotics programs, providing R&D teams with actionable hardware selection and integration guidance.

Jetson AI Agent Mass‑Deployment Practice — Jin Liancheng, Project Manager, RealTimes

Drawing on multiple deployed projects, Jin Liancheng shared hands‑on experience in integrating AI agents on the Jetson platform. He covered system integration, software‑stack deployment, troubleshooting, and project‑delivery management, and explained how to balance algorithm requirements, hardware performance, and project schedules—offering a complete reference for moving AI‑agent projects from prototype to small‑batch production.

New Perception Technologies for Embodied Intelligence: Next‑Generation Vision Solutions — Li Bin, Chief Market Ecosystem Officer, Shenzhen Senyun Intelligent

Li Bin addressed the robot perception layer, introducing next‑generation vision solutions for embodied robots. He explained multi‑camera fusion, high‑dynamic‑range imaging, and environmental perception to improve understanding of complex scenes, shared integration approaches between vision sensors and Jetson‑based embodied brains, and discussed how perception optimization enhances autonomous interaction.

Highlight: Hands‑On Exhibition Zone for Jetson Embodied Brains

The venue included a live hardware showcase where core Jetson‑based embodied‑brain prototypes were displayed. Engineers could examine the full hardware systems up close; RealTimes technical staff explained interfaces, compute configuration, and thermal design on site, and demonstrated multi‑sensor interfacing and on‑device AI inference. Many attendees brought their own project requirements and held one‑on‑one discussions with the technical team on hardware customization and project adaptation.

Networking Breaks: Face‑to‑Face Industry Exchange

The refreshment breaks provided a relaxed setting for exchange. Technical leaders from robot OEMs, algorithm companies, research institutes, and component suppliers networked freely, discussing embodied‑intelligence collaboration, compute‑platform selection, joint software‑hardware development, and supply‑chain support. Beyond a technical‑sharing platform, the seminar served as an industry window connecting upstream and downstream partners as well as research and application resources.

The Shenzhen NVIDIA Jetson Embodied‑Brain Mass‑Deployment Practice Seminar concluded successfully. Scaling embodied intelligence into commercial use requires continuous iteration of the underlying compute foundation and, more importantly, collaborative engineering by ecosystem partners to solve mass‑production challenges.

Moving forward, RealTimes will continue to deepen its work in the NVIDIA Jetson ecosystem, refine production‑grade embodied‑brain solutions, and host ongoing technical exchanges—partnering with the industry to move embodied intelligence from the lab into real‑world scenarios and accelerate robotics deployment.