- Hyundai Motor Group’s Robotic LAB announces on-device AI chip for robots' “Edge Brain” utilizing its proprietary software, to be produced by DEEPX under a three-year partnership
- Integration of Robotics LAB’s software capabilities and DEEPX's semiconductor expertise targeting market leadership
- "Robotization of Space" solutions to be deployed across factories, buildings, logistics, and mobility sectors
- Group aims to advance social value through advanced robotics and physical AI systems, addressing labor shortages and industrial challenges
Hyundai Motor Group's Robotics LAB revealed a pivotal accomplishment in physical artificial intelligence[1] (AI) technology at the inaugural CES Foundry 2026, showcasing the outcomes of its strategic three-year partnership with AI semiconductor specialist DEEPX. The announcement marks the start of mass production for an “edge brain” on-device AI[2] chip, co-developed by the Robotics LAB and DEEPX, with DEEPX producing the chip that integrates the Robotics LAB’s proprietary software. This collaboration is critical for empowering robots to act autonomously without external connectivity.
As part of the newly established CES Foundry — a dedicated exhibition at CES 2026 spotlighting innovations in AI, blockchain, and quantum technologies — the Group’s milestone contribution exemplifies how these groundbreaking advancements intersect with real-world applications.
"Hyundai Motor Group’s Robotics LAB is not just focused on building robots but on creating a sustainable robotics ecosystem. Our aim is to deliver real value at the final touchpoint of the Physical AI pipeline, where users directly interact with the technology, and to make efficient, smart, and low-power robots accessible to a wider audience." – Dong Jin Hyun, Vice President and Head of Hyundai Motor Group's Robotics LAB.
What Is the “Edge Brain” On-Device AI chip?
Following three years of collaboration with DEEPX initiated by a 2023 memorandum of understanding (MOU), the two companies co-developed the AI chip, with the Robotics LAB designing the control software applied to the chip and DEEPX leading its production.
The ultra-low power chip consumes less than 5 watts of power during operation and serves as an "edge AI[3] brain", enabling robots to recognize their surroundings in real time and make autonomous decisions and controls without cloud or network connectivity. This core technology for physical AI implementation provides the foundation for robots to operate reliably in everyday life and industrial settings.
This technology has already been integrated into the Robotics LAB’s technologies, including Facey, a cutting-edge facial recognition technology currently under development, and DAL-e Delivery, an advanced delivery robot currently in the demonstration and validation phase. Going forward, it will be applied sequentially to the Robotics LAB’s AI security solutions and next-generation mobile robots.
How Does This Development Align with the Group’s Robotics Vision?
The future roadmap of the Group's Robotics LAB focuses on going beyond individual robot products to providing “Robotics Total Solution (RTS),” integrating hardware and software. The development of the on-device AI chip exemplifies this approach, as it merges the Robotics LAB’s control software expertise with DEEPX’s AI semiconductor capabilities.
This collaboration has also enabled the Group to secure optimized solutions in advance for future mass-produced robots, while simultaneously enhancing flexibility and stability within its supply chain. Ensuring such robots can operate effectively in real-world environments is crucial to addressing societal challenges such as aging populations, industrial safety concerns, and labor shortages. To this end, the Group anticipates that the on-device AI chip developed through this initiative will enable the establishment of stable physical AI infrastructure to help combat such challenges.
Leveraging decades of experience in automotive value chains, the Group is dedicated to achieving the stable mass production of robots while strengthening partnerships with the domestic battery industry. Furthermore, by expanding industrial applications such as pilot projects at airports and hospitals, the Group is striving to broaden the domestic robotics ecosystem.
Unlocking Next-Generation Robotics Through Partnerships
Recent advancements supported by partners further highlight the Group’s leadership in robotics. In December 2025, the Group launched MobED (Mobile Eccentric Droid), its autonomous mobile robot platform, at iREX. This groundbreaking platform is the first autonomous driving robot platform prepared for mass production in the global automotive sector and received the prestigious Best of Innovation Award in Robotics at CES 2026.
The latest partnership with DEEPX underscores the Group’s commitment to strengthening its optimized RTS through early validation, advancing convergence technologies while reinforcing a stable supply chain.
[1]Physical AI: Artificial intelligence that operates in the physical world. It enables robots or devices to make autonomous decisions and act in real environments rather than digital spaces.
[2]On-device AI: A technology where artificial intelligence operates on the device itself rather than cloud servers. Enables real-time decisions without network connectivity.
[3]Edge AI: Artificial intelligence processed immediately at the edge where data is generated, rather than on central servers. Offers advantages in fast response time and security.
