Qualcomm held its Dragonwing IoT Day event in late October, focusing on semiconductor solutions for the IoT and physical AI markets, including robotics, and Qualcomm’s related software and support ecosystem. Counterpoint’s Greg Basich, Associate Director, Automotive, spoke with Anshuman Saxena, VP and General Manager, Robotics and ADAS, at Qualcomm about what it brings from its experience in the automotive industry to the robotics market and what differentiates the robotics market from the auto industry.

The Interview

Key Takeaways from the Discussion

Qualcomm Developing a Hardware, Software, and Data Platform for the Robotics Market

• Qualcomm is developing a software and data platform for the robotics market that is integrated with its line of Dragonwing IQ processors.
• Developing effective AI models, such as VLA models, for robotics applications requires training data, which is in short supply.
• Qualcomm is taking its solutions for the automotive industry, and what it has learned from that segment, to the robotics industry while simultaneously developing new solutions designed to address the robotics industry’s broader scope.

Acquisitions Are Also Key to Qualcomm’s Efforts

• As part of Qualcomm’s development efforts, during this interview Saxena discussed the company’s acquisition of PickNik, which manages the open-source MoveIt robotics manipulation framework. 
• Qualcomm intends to further integrate MoveIt Pro, the commercial version of MoveIt, with its Dragonwing robotics platform (though it’s important to note that the open-source version of MoveIt will remain so). 
• The company is also integrating its solutions with Arduino’s products (Qualcomm announced its acquisition of Arduino in October 2025) to reach developers using Arduino for product development and prototyping.

Analyst Take

• Qualcomm has been in the automotive market for many years, providing cellular connectivity solutions to the industry, and the advanced driver assistance system (ADAS) market segment for a decade. The company has focused on developing a full data “flywheel,” including solutions and practices for curating data, annotating it, and training models. 
• In the robotics market, the environments are often unstructured and scenarios are complex, since “robot” is a general term for a machine that senses its environment and independently performs tasks. 
• Scenarios vary widely; for example, industrial robots used in manufacturing, autonomous mobile robots (AMRs) for cleaning or warehouse and logistics tasks, and humanoid robots, which are used for service and guidance and in manufacturing and warehouse logistics roles in pilot programs, among other scenarios.
• As a result, companies in the robotics segments are gathering training data they can use to train more capable AI models, such as Vision-Language-Action (VLA) models. 
• Qualcomm is using its infrastructure and what it has learned in the ADAS market for the robotics market. The company has developed a new software stack that works with its Dragonwing IQ line of processors.
• Although Qualcomm is bringing its background and technology for the automotive industry to the robotics market, the company is developing its own foundational hardware and software platform designed to accommodate the wide range of use cases in the robotics segment.