The first question for any robotics startup is whether it builds the robot. Noitom Robotics, a Beijing-based spinout from motion-capture leader Beijing Noitom Technology, has a different answer. It sells the hardware and data that other companies need to build theirs. In its first eight months of operation, the company reported signed orders worth tens of millions of yuan, a five-fold increase over its prior year [36Kr, Aug 2025]. For founder Dai Ruoli, a repeat founder and CTO of the parent company, the bet is that the bottleneck for embodied AI isn't the robot body, but the high-fidelity data needed to train it.
The wedge is motion-capture hardware
Noitom Robotics's initial wedge is a high-precision, low-latency motion-capture system designed for robot teleoperation. Under the banner of 'Project T,' the company sells this hardware to robot original equipment manufacturers (OEMs). The setup allows a human operator to control a robot in real time, generating the kind of nuanced, real-world movement data that is otherwise scarce and expensive to collect. The company claims this approach can reduce robot deployment times from months to weeks for specific tasks [LogiShift, Dec 2025]. The business model is straightforward: sell the gear, then sell the services to help customers deploy the resulting data.
A planned pivot to data licensing
While Project T focuses on hardware and services, the company's longer-term strategy is codenamed 'Project D.' This initiative aims to build a 'data collection factory' that will produce and sell curated, licensed datasets for embodied intelligence [36Kr]. The move from selling shovels to selling the gold, so to speak, represents a potential shift to a more scalable, higher-margin revenue stream. The company emphasizes that these datasets will come with commercially friendly licensing terms, a direct response to the restrictive or unclear licensing that often plagues public robotics datasets [Robotics Center, 2026].
The founder's track record as an asset
The company leans heavily on the technical and commercial credibility of founder Dai Ruoli. His prior venture, Beijing Noitom Technology, captured an estimated 70% of the global motion-capture market, shipping over 15,000 systems to more than 50 countries [Noitom Robotics]. This track record in a closely adjacent field is a significant asset. It provides Noitom Robotics with an immediate technological foundation and, presumably, a supply chain and manufacturing edge. Dai's background also includes a stint in sales, a practical skill that likely informs the company's go-to-market approach [Spotify for Creators]. The founding team is described as veterans across robotics, computer vision, and high-precision tracking [Noitom Robotics].
Where the model faces pressure
Noitom Robotics's model is not without its counter-bets. The company operates in a crowded field of established motion-capture providers, many of whom could theoretically pivot to serve the robotics market. The competitive set is long and includes global players like Vicon, Xsens, and Rokoko, as well as a host of Chinese specialists. The company's reported traction is impressive but lacks specific, named customer logos, which makes it harder to gauge true market penetration. Furthermore, the success of Project D hinges on a market that is still forming. Robot makers must first be convinced that buying curated data is more efficient than generating it themselves, a calculation that depends on the quality, specificity, and price of Noitom's offerings.
The company's early traction signals are concentrated in a few key areas:
- Revenue growth. Reported signed orders grew five-fold in the first eight months of 2025 compared to the prior full year [36Kr, Aug 2025].
- Investor validation. The company has secured Pre-A and Pre-A+ funding rounds from a slate of notable Chinese venture firms, including Qiming Venture Partners and Alpha Community [Pandaily, Dec 2025].
- Partnership development. It has launched a one-stop data acquisition package in partnership with other robotics firms like PND Robotics and Inspire Robots [Interesting Engineering].
The ideal customer profile here is clear: a humanoid or industrial robot OEM that is scaling production and needs to generate vast amounts of training data quickly, but lacks the internal motion-capture expertise. They are likely mid-sized companies racing to catch up to larger players, for whom building this capability in-house would be a costly distraction. The realistic competitive set splits into two camps. First, the incumbent motion-capture giants like Vicon and Xsens, who have the technology but may lack the specific robotics integration focus. Second, a wave of AI-first data companies that are building synthetic or real-world datasets, but without the integrated hardware play. Noitom's bet is that its combined hardware-and-data offering, backed by Dai's domain authority, carves out a defensible niche between them.
Sources
- [36Kr, Aug 2025] Exclusive Report from Intelligence Emergence | https://eu.36kr.com/en/p/3478931911252355
- [LogiShift, Dec 2025] Project T enables robot deployment time to drop from months to weeks | https://logishift.com
- [Robotics Center, 2026] Project D aims to provide commercially friendly licensing | https://roboticscenter.com
- [Noitom Robotics] Inside Noitom Robotics: Experts in Embodied AI Systems | https://noitomrobotics.com/about/
- [Spotify for Creators] Dai By Dai Podcast | https://creators.spotify.com/pod/profile/daibydai
- [Pandaily, Dec 2025] Noitom Robotics Raises Pre-A+ Round Led by Qiming Venture Partners | https://pandaily.com/noitom-robotics-raises-pre-a-round-led-by-qiming-venture-partners-positioning-itself-as-a-robot-company-that-doesn-t-build-robots
- [Interesting Engineering] Partnered with PND Robotics and Inspire Robots | https://interestingengineering.com