At CES, Qualcomm Technologies, Inc. unveiled a comprehensive suite of robotics technologies designed to power Physical AI across a wide range of applications—from household robots to full-size humanoids. Central to the announcement is the new Qualcomm Dragonwing™ IQ10 Series, a high-performance robotics processor tailored for industrial autonomous mobile robots (AMRs) and advanced humanoid systems. This latest addition to Qualcomm’s robotics roadmap builds on its expertise in edge AI, energy efficiency, and low-latency, safety-critical computing. The Dragonwing IQ10 Series is engineered to serve as the “brain” of intelligent robots, delivering powerful, efficient, and scalable processing for complex tasks. It supports advanced capabilities such as real-time perception, motion planning, and end-to-end AI models like Visual Language Action (VLA) and Visual Language Models (VLM), enabling robots to perform generalized manipulation and interact naturally with humans and their environments. Nakul Duggal, Executive Vice President and Group General Manager of Automotive, Industrial & Embedded IoT and Robotics at Qualcomm Technologies, emphasized the company’s focus on making intelligent machines reliable, safe, and ready for real-world deployment. “By leveraging our foundation in low-latency, safety-grade, high-performance systems—from sensing and perception to planning and action—we’re moving physical AI beyond the lab and into practical, large-scale use,” Duggal said. The new architecture is already being adopted by leading robotics companies. The Dragonwing IQ9 Series powers VinMotion’s Motion 2 humanoid, which will be on display at Qualcomm’s booth (No. 5001) at CES. Booster’s K1 Geek robot also highlights Qualcomm’s leadership in edge AI, demonstrating the company’s commitment to enabling developers and organizations to build next-generation robotic systems. Qualcomm is also collaborating with Kuka Robotics on its next-generation robotics platform, further expanding its footprint in industrial automation. The company is showcasing a range of tools and solutions at CES, including Advantech’s commercial robotics development kit, which accelerates multi-application development and deployment. Additionally, the booth features teleoperation tooling and an AI data flywheel that supports continuous learning, training, and skill deployment across different robot types. The full-stack robotics architecture combines heterogeneous edge computing, mixed-criticality systems, machine learning operations, and a growing partner ecosystem. Supported by developer tools, this end-to-end approach allows robots to reason, adapt, and operate reliably in dynamic, real-world settings. Qualcomm’s vision is to make physical AI accessible and scalable across industries. By integrating powerful AI, efficient computing, and robust connectivity, the company is helping bridge the gap between robotics prototypes and deployable, production-ready systems. The result is faster innovation, improved productivity, and the potential to transform how humans and machines work together.