Qualcomm said its architecture can enable capabilities in service robots, industrial mobile robots, humanoids, and more. | Source: Qualcomm
Qualcomm Technologies Inc. this week introduced its next-generation robotics architecture stack, which integrates hardware, software, and compound artificial intelligence. At CES, the company also unveiled its latest high-performance robotics processor for industrial mobile robots and full-size humanoids, the Qualcomm Dragonwing IQ10 Series.
“As pioneers in energy efficient, high–performance physical AI systems, we know what it takes to make even the most complex robotics systems perform reliably, safely, and at scale,” said Nakul Duggal, executive vice president and group general manager for automotive, industrial and embedded IoT and robotics at Qualcomm.
“By building on our strong foundational low-latency, safety-grade, high-performance technologies, ranging from sensing, perception to planning and action, we’re redefining what’s possible with physical AI by moving intelligent machines out of the labs and into real-world environments,” he asserted.
Qualcomm has more than 40 years of experience in developing technology. Its portfolio includes edge AI; high-performance, low-power computing; and connectivity. The San Diego, Calif.-based company offers its Snapdragon platform for consumers, while its Dragonwing line is geared towards businesses and industries.
Robotics architecture moves from concept to development
Figure is targeting a range of applications in industrial and home environments with its humanoid. | Source: Figure AI
Today, the Dragonwing industrial processor roadmap powers an assortment of general-purpose form factors, including humanoid robots from Booster, VinMotion, and other global robotics providers.
“Figure’s mission is to develop general-purpose humanoid robots powered by advanced AI to eliminate unsafe and undesirable jobs, boost productivity across industries, and create economic abundance that enables happier, more purposeful lives for humanity,” stated Brett Adcock, the founder and CEO of Figure AI. “Qualcomm Technologies’ platform, with its combination of exceptional compute capabilities and energy efficiency, is a valuable building block in enabling Figure to turn our vision into reality.”
Qualcomm added that new architecture supports advanced perception and motion planning with “end-to-end” AI models such as vision-language-action models (VLAs) and vision-language models (VLMs), enabling generalized manipulation capabilities and human-robot interaction. The company said Dragonwing IQ10 helps it “take a significant step toward practical, real‑world deployment across industrial applications.”
In addition, Qualcomm said it is in discussions with KUKA Robotics for its next-generation robotic system.
SITE AD for the 2026 Robotics Summit save the date.
Qualcomm builds an integrated, supported stack
The general-purpose robotics architecture and the Dragonwing IQ10 combine heterogeneous edge computing, edge AI, mixed-criticality systems, software, machine learning operations, and an AI “data flywheel,” Qualcomm said.
The company claimed that its approach enables robots to easily reason and adapt to the spatial and temporal environment. It is optimized to scale across various robotic form factors with industrial-grade reliability, according to Qualcomm.
In addition, Qualcomm’s technologies are supported by a growing partner ecosystem and complemented by a suite of developer tools. This collaborative network can accelerate robotics development and deployment, solving the last-mile challenges and enabling faster, more scalable innovation across industries, it said.
VinMotion’s Motion 2 humanoid, powered by the Dragonwing IQ9 Series, is on display at Booth 5001 during CES. Qualcomm’s booth also features Booster’s K1 Geek, which it said demonstrates Qualcomm’s leadership in edge AI and commitment to advancing physical AI.
Qualcomm is also showing Advantech’s commercially available robotics development kit for rapid, multi‑application development and deployment. Separately, the booth features an in-depth look into teleoperation tooling and an AI data flywheel for collection, training, and deployment to continuously add new skills across robotic form factors.