INFIFORCE Raises Nearly 1 Billion RMB in Series A: AtomBrain Turns "The Robot's Own Eyes" into a Data Moat
On August 14, embodied intelligence company INFIFORCE (原力无限) announced the completion of its Series A and A+ financing rounds, totaling nearly 1 billion RMB. The round was led by Dunhong Asset and a leading state-owned investment platform, with participation from industry capital including Zhejiang University Sci-Tech Innovation Group, Yandu Guokong, and Lishui Municipal State-owned Capital, plus continued backing from existing investor Creation Venture Partners (CCV). Proceeds will focus on three areas: the AtomBrain causal world model, the full-stack AI infrastructure DataGrid, and scaled delivery of multi-form robot hardware.
Breaking Down the Round: Four Types of Capital Betting Together
The investor mix is the most noteworthy signal of this financing:
- Dunhong Asset: hard-tech institution focused on advanced manufacturing and embodied AI
- Leading state-owned platform: strategic long-term capital representing local government industrial priorities
- Zhejiang University Sci-Tech Innovation Group: industry-academia platform bringing university resources
- Yandu Guokong and Lishui state capital: local industrial capital from Yancheng (Jiangsu) and Lishui (Zhejiang)
- Creation Venture Partners (CCV): existing investor doubling down
- AtomBrain: a causal world model unifying visual observations, language instructions, action trajectories, and environmental feedback (force, position, velocity) in a single framework
- AIM world model: third generation, reportedly reaching 93.1% task success rate on the RoboTwin 2.0 benchmark
- Early Atom models: 97.0% task success on the LIBERO benchmark
- Founder & CEO Bai Huiyuan: former Alibaba VP; gave the talk "Awakening the Force: The Rise of Embodied Intelligence" at TEDxXuhui
- CTO Wang Yizhou: UC Berkeley PhD, among the first core members of Baidu IDL's autonomous driving team and a core member of NVIDIA DriveAV
- Chief Scientist Chen Jiayu: Assistant Professor at the University of Hong Kong, CMU postdoc
- Core R&D team: over 75% hold master's or doctoral degrees, with members from Alibaba, Huawei, Baidu, NVIDIA, and other leading companies
- IP: 100+ cumulative patents and software copyrights
- Standards: In July 2026, the GB/Z "Technical Specification for Crowdsourced Data Collection and Management of Embodied Intelligence," initiated and led by INFIFORCE, was approved for立项 — the industry's first national-level embodied intelligence data specification
Financial capital + strategic state capital + local industrial capital + academia — a rare combination in the embodied AI space. It suggests INFIFORCE is positioned not just as a technology play, but as a "region-level brain company" embedded in local government's embodied intelligence industry plans.
AtomBrain: An Engineering Bet on Ego-Centric Data
INFIFORCE is betting on a route no one has fully proven yet: Ego-Centric (first-person perspective) data.
Mainstream embodied models rely on two data types: internet videos (large volume but noisy human-behavior footage) or teleoperation collection (high quality but extremely expensive and hard to scale). INFIFORCE chose a third path: let robots see with their own eyes and act with their own hands, collecting first-person interaction data from their own perspective.
Analogy: human infants don't learn object manipulation by watching videos — they look with their own eyes, touch with their hands, feel force feedback, and learn by trial and error. The core advantage of ego data: the robot uses information it actually perceives, with no "third-person" annotation bias.
Around this route, INFIFORCE built the Atom series of native embodied foundation models:
Supporting model training is the company's end-to-end data and training system DataGrid, covering UMI handheld grippers, ego-collection devices, and real-robot teleoperation — connecting the pipeline from "real-world experience → model capability."
Commercial Delivery: 100 Scenarios Across 30 Cities
Models alone aren't enough — the real threshold in embodied AI is validation in real scenarios. INFIFORCE disclosed that, based on a unified AtomBrain, it has built a full product matrix spanning specialized, general-purpose, and humanoid categories: AstroDroid wheeled humanoid, UltraDroid multi-form general-purpose, the "Little Atom" bipedal, and the FORCE specialized series.
Key number: commercialization in 30+ cities and 100+ real-world scenarios nationwide, covering commercial services, warehousing/logistics, and industrial manufacturing.
100 scenarios is not a small figure. Even at 1–2 robots per scenario, that's a 100–200 unit real deployment — enough to spin the ego data flywheel (more deployments → more data → stronger models → better deployment performance).
Team Background: Alibaba, Berkeley, NVIDIA DriveAV
Industry Signal: From "Whose Body Is Better" to "Whose Brain Is Smarter"
A key line from INFIFORCE's announcement: "The core of competition in embodied intelligence has shifted from 'who can build the robot body' to 'who can keep making robots smarter.'"
Over the past two years, embodied AI funding has centered on hardware companies — humanoids, robotic arms, sensors. That logic is now being questioned: good hardware does not equal good intelligence. Tesla Optimus hardware is already quite advanced, yet its intelligent task performance in factories remains a bottleneck.
INFIFORCE's bet is brains + data. If robot hardware companies "build bodies," INFIFORCE "builds brains plus the real-machine data loop needed to train them."
Four validation points to watch over the next 6–12 months: 1) AtomBrain's published results on third-party benchmarks (RLBench, RoboTwin 2.0, Calvin); 2) how many of the 100 scenarios achieve stable operation for 6+ months; 3) customer repurchase and renewal rates across the 30 cities; 4) whether real "brain-across-body transfer" cases emerge (the same AtomBrain running on different manufacturers' robot hardware).
Sources: INFIFORCE announcement, Blue Shark Hard Tech 8-14 financing review, Tencent embodied AI daily 8-15, 10jqka 8-15 report.