年表作成・共有サービス - pastport

CES 2026: The "Physical AI" Revolution - How Sensors and Silicon are Bringing Robots to Life
Publication Date: January 12, 2026
Keywords: CES 2026, Physical AI, Robot Sensors, LiDAR, STMicroelectronics, Lomotive, Arbe, NVIDIA
The 2026 Consumer Electronics Show (CES) has drawn to a close, leaving no doubt about the future of technology. While the past two years were dominated by Large Language Models (LLMs) chatting on screens, CES 2026 was all about "Physical AI".
The focus has decisively shifted from virtual conversations to real-world interaction. As NVIDIA CEO Jensen Huang declared, we have reached the "ChatGPT moment for robotics." But for robots to function in our complex world, they need more than just a brain—they need superhuman senses.
This article dives into the unsung heroes of CES 2026: the advanced sensors, chips, and LiDAR systems that are giving machines the ability to see, touch, and understand the physical universe.

1. The Rise of the "Digital Eye": High-Resolution LiDAR and ToF
To navigate a world built for humans, robots require high-definition spatial awareness. Several key players unveiled breakthroughs that move us beyond simple obstacle detection.
STMicroelectronics: Bringing LiDAR to Your Pocket
In a move that could democratize high-end robotics, STMicroelectronics introduced the VL53L9, a "direct ToF (Time-of-Flight) 3D LiDAR" sensor.
● The Innovation: It features up to 2,300 measurement zones, a massive leap from traditional single-point sensors.
● Why it Matters: This resolution allows robots to recognize object contours and edges, not just distance. Combined with its "dual-scan" illumination, it can detect small objects and complex shapes, effectively turning a simple sensor into a low-power 3D imaging node.
Seyond (RoboSense): The "Pure Solid-State" Breakthrough
Seyond stole the spotlight with its Hummingbird D1, a pure solid-state LiDAR that won the "Best-in-Show" award.
● The Innovation: Unlike bulky spinning units, this chip-based LiDAR has no moving parts. It allows for a sleek, invisible integration into car roofs or robot chassis.
● The Impact: It signals a shift toward mass-producible, maintenance-free sensors that are essential for consumer robots and autonomous vehicles.

2. Silicon Photonics: The Infrastructure Play
Beyond the big names, a quiet semiconductor revolution was on display. Investors like Bill Gates and Amazon are betting big on a new class of "optical semiconductors."
Lomotive Labs: Programmable Light
Lomotive introduced a Programmable Optical Chip that could change how we build sensors.
● The Concept: Instead of using mechanical mirrors, this chip controls light beams software-defined. One chip can act as multiple "virtual sensors."
● The Advantage: It’s smaller, cheaper, and more reliable than mechanical systems. This "software-defined sensing" allows developers to tweak a robot's "vision" in real-time without changing hardware, perfectly aligning with the needs of Physical AI.
Arbe Robotics & NVIDIA: Seeing Through the Storm
For autonomous driving and outdoor robots, seeing in the dark or bad weather is non-negotiable. Arbe partnered with NVIDIA to showcase their Ultra-HD 4D Imaging Radar.
● The Power: Capable of detecting objects over 300 meters away with extreme precision (2,304 channels!).
● The Result: This system acts as a "super sense," allowing a robot or car to "see" through fog, dust, or rain, providing the redundancy needed for true Level 4 autonomy.

3. The "Feeling" Factor: Safety and Tactile Sensors
Physical AI isn't just about sight; it's about safe interaction. New technologies ensure robots can work with humans, not just alongside them.
RoboSense: Safety Airy
RoboSense launched the world's first 3D Safety LiDAR.
● The Shift: Moving from 2D lines to near-hemispherical 3D protection.
● The Benefit: It can detect if a hand or a small object enters a dangerous zone around industrial machinery. This is critical for collaborative robots (cobots) in factories or service robots in homes.

1. ##帕西尼 (PASSTINY): The Delicate Touch

Demonstrating at CES,帕西尼's technology allowed a robot to make ice cream.
● The Tech: High-precision tactile sensors that mimic human skin.
● The Significance: This allows robots to handle fragile objects (like eggs or delicate ingredients) with the right amount of grip force, opening doors for robots in kitchens and healthcare.

4. The Road Ahead: Key Trends for 2026
Based on the innovations at CES 2026, the trajectory for robotic sensing is clear:
1. Sensor Fusion is King: No single sensor is enough. The winning platforms (like those using NVIDIA's DRIVE Orin chips) will seamlessly blend LiDAR, Radar, and Camera data.
2. Chipization (The "Camera" Moment): Just as CMOS sensors enabled the smartphone camera boom, the "chipization" of LiDAR and Radar (like the ST and Lomotive products) will drive costs down and reliability up.
3. From Lab to Living Room: With products like intelligent lawn mowers (Navimow) and pool cleaners (Ecovacs) utilizing these advanced sensors, Physical AI is moving out of research labs and into commercial profitability.
Conclusion:
CES 2026 proved that the era of the "dumb robot" is over. Thanks to breakthroughs in optical chips, high-resolution LiDAR, and tactile sensing, the robots of tomorrow will be able to perceive the world with a clarity and safety that was previously science fiction. The future is not just intelligent; it is physical.
https://tofsensors.com