Will the 6D force sensor equipped with AI sail into a blue ocean market?
Release Time:
2025-11-05
Where does AI's "sensing" come closest to that subtle, almost spiritual feeling? Recently, Apple published a new AI paper that transforms raw data collected by LiDAR sensors into 3D maps, marking a significant step forward in shifting sensor-derived information from mere raw data toward fully immersive, three-dimensional representations. While we’re still quite some way from capturing that "spiritual" essence, this research nonetheless inspires us to focus more intently on one critical aspect of human-machine interaction: how we acquire and process information.
Equipped with "First Brain" sensors, it can be smarter.
Two distant relatives of "AlphaGo" have also become popular: one is Sogou's Chinese AI assistant, "Wangzai," which showcased "lip-reading recognition" at the Internet Conference; the other is "Alphaba," the driverless bus that has already been put into operation in Shenzhen.
The former breaks conventional thinking by ingeniously replacing the traditional semantic recognition sensor with an optical one, using image-captured data to interpret and facilitate language-based communication. Meanwhile, the latter features strikingly eye-catching, flat-round sensors embedded in the vehicle body—thanks to these, the "Alphaba" can perceive its surroundings and effortlessly avoid obstacles.
Humans acquire 80% of their information through their eyes; similarly, in the process of AI capturing data, visual sensors play a remarkably important role—currently, the two main methods are radar and video. Compared to radar, video offers a more holistic view, making it less susceptible to interference, while radar excels at creating 3D models of the surrounding environment, enabling it to capture richer depth information than even advanced photographic cameras.
“Currently, the obstacle-sensing devices in use include microwave radar, ultrasonic radar, and other technologies—but there’s also a method that relies on capturing video images,” explained Bi Chao, a technical staff member at the Beijing Intelligent Connected Vehicle Industry Innovation Center. He added that these sensing devices can be installed either on vehicles or along roadsides, with the critical requirement of achieving seamless coverage, “much like the relationship between a mobile phone and its base station, ensuring a stable and uninterrupted signal.”
"Smooth communication is the foundation; accurate judgment is the key. 'Automobiles require instantaneous, split-second decisions and actions—so calculating the likelihood of collisions between vehicles, between vehicles and roads, and between vehicles and pedestrians must happen incredibly fast,' says Bi Chao. He adds that processes like information acquisition, transmission, computation, and feedback need to flow seamlessly in one continuous motion."
Specifically, before braking or triggering the alarm, numerous calculations take place: The radar captures three-dimensional point data, identifies obstacles, compares the obstacle’s characteristics across two consecutive frequency scans—determining whether it’s static or dynamic. For dynamic obstacles, the system calculates their speed of movement and, combined with the autonomous vehicle’s own positional information, computes the safe distance required to avoid the obstacle, ultimately deciding on the appropriate response.
How can we speed things up? The industry is experimenting with moving the "pre-processing" part of data centers closer to the sensor end. "We’ve been testing the multi-dimensional force sensor," says Liu Zhengzhong, a senior engineer at Shengzhe Technology. "Traditionally, sensors only had the single function of collecting signals—but these signals had to travel back and forth between the sensor and the processing unit, consuming both time and energy. If the six-axis force sensor could filter out some of the unnecessary information upfront, the entire process would become significantly more streamlined."
"In other words, the previous sensors were strictly tied to our sensory organs—hands, eyes, ears, nose, and so on—while now they’re equipped with tiny 'pre-brains.' 'This is particularly effective for video-based sensors, since video data volumes are simply enormous,' said Liu Zhengzhong."
“We’ve already studied cameras thoroughly, but radar still requires further research. We also need to explore more deeply: What kind of sensors can detect those subtle changes that humans perceive?” Philip said. Meanwhile, the AI market is rapidly evolving in its demand for sensors, creating a strong need for the development and adoption of next-generation sensor technologies.
Get a Free Consultancy
We will contact you within one working day. Please pay attention to your email.
Nanjing Bio-inspired Intelligence Technology Co., Ltd.
南京神源生智能科技有限公司
Contact Us
Webchat
Copyright © 2025 Nanjing Bio-inspired Intelligence Technology Co., Ltd.