Sensing, imaging, and RF: the three pillar technologies serving “physical AI”

The combination of RF, imaging, and sensing markets generated almost $100 billion in revenue, in 2025 and will grow at a 4.4% CAGR_25-31.

As all three industries mature, players are building and consolidating expertise to address better challenges related to product development, volume scalability, and profitability across automotive, consumer, AI, and robotics-related markets.

Today’s snapshot, written by Cyril Buey, Anas Chalak, and Clyde Midelet, Market & Technology Analysts at Yole Group, comes from the broad collection of “Status of” analyses, including Status of the RF Industry 2026, Status of the CMOS Image Sensor Industry 2026, Status of the MEMS Industry 2026, and more.

Yole Group also extends its thanks to the invited contributors from Bosch Sensortec, Infineon Technologies, Sonair, and STMicroelectronics.

Enjoy reading this new Yole Group’s strategic insight!

For decades, machine perception meant a camera bolted onto a fixed-position robot arm, reading a barcode or checking a weld seam against a narrow set of rules. Today’s perception stack including imaging, sensing, and radio-frequency (RF) technologies, has become the sensory nervous system of a new generation of machines: humanoid robots, autonomous mobile platforms, drones, and vehicles that must operate in unstructured, unpredictable human environments. What has changed is not just the hardware, but what sits behind it. The rise of Generative AI and large multimodal models has transformed perception from a narrow, task-specific pipeline into a general-purpose understanding layer that can reason about a scene rather than merely detect it. This is the essence of “Physical AI“: intelligence that doesn’t just generate text or images, but perceives, reasons, and acts in the physical world thanks to sensor fusion, connectivity, and high computing capabilities.