SEOUL, South Korea, Dec. 3, 2025 /PRNewswire/ — DEEPX, an ultra-low-power on-device AI semiconductor company, is unveiling a next-generation robot intelligence platform co-developed with Hyundai Motor and Kia’s Robotics LAB. The controller platform has now progressed into the commercial validation stage for real-world deployment, in preparation for future mass-production, marking a key milestone toward the commercialization of Physical AI — enabling robots to operate autonomously without reliance on cloud connectivity.
Hyundai Motor, Kia’s Robotics LAB and DEEPX Begin Commercialization for Next-Generation On-Device AI Robot Platform
DEEPX’s DX-M1 NPU has been integral to the robot development roadmaps of Hyundai Motor and Kia’s Robotics LAB since 2023 and offers sub-5W power consumption, high-performance inference, and low latency. These properties make it suitable for indoor and outdoor service robots operating under strict power and thermal constraints.
In 2024, the teams developed a new controller architecture combining the DX-M1 with a dual wide- and narrow-angle ISP camera system and the LAB’s proprietary vision AI technology. This design allows robots to function reliably in network-restricted environments — including underground facilities, transportation hubs, and logistics centers — without depending on cloud services.
The DX-M1 also powers the LAB’s facial recognition system, Facey. Leveraging this foundation, the LAB’s DAL-e Delivery robot has already demonstrated recipient authentication, user identification, and guided interaction capabilities, paving the way for more advanced service-robot functions.
The next-generation robot intelligence platform from DEEPX, Hyundai Motor and the LAB will be presented at major global industry events starting in December and showcased publicly at CES 2026 in Las Vegas.
Looking ahead, DEEPX will continue working with Hyundai Motor and Kia’s Robotics LAB to expand their collaboration, accelerating the development and deployment of Physical AI systems across manufacturing, logistics, mobility, and smart-city applications.
