--- title: "He watched therapists hand-move patients' legs — then built China's rehab robots and its first mass-produced humanoid" date: 2026-10-18 category: Builders site: NeuroAI canonical: https://neuroai.site/a/na-people-gu-jie-fourier language: en --- # He watched therapists hand-move patients' legs — then built China's rehab robots and its first mass-produced humanoid > Gu Jie founded Fourier Intelligence (傅利叶智能) after the 2008 earthquake showed him rehab robots cost a fortune, then pushed into humanoids. In 2008, after the Wenchuan earthquake, Chinese hospitals were flooded with patients who needed months of repetitive rehabilitation — and the imported robots that could help cost between **RMB 5 million and 6 million (≈ US$695K–835K / HK$5.4M–6.5M)** apiece. A young engineer named Gu Jie (顾捷) looked at that gap and decided it was the problem he wanted to spend his life on. He did not set out to build a humanoid (人形机器人). He set out to build one a hospital could actually afford. ## A rehab robot before a celebrity robot Gu Jie graduated from Shanghai Jiao Tong University's mechanical engineering department in 2003 and worked at National Instruments before starting his own path in robotics. The founding moment for Fourier Intelligence — **傅利叶智能** — came in 2015, in a small office in Zhangjiang, Shanghai's science park. The name is borrowed from the French mathematician, a nod to signal and frequency, the language of motion and force. The first years were deliberately unglamorous. Gu's team embedded themselves in hospital rehab departments, watching therapists physically lift and rotate patients' limbs for hours. The core engineering problem turned out to be force feedback: sensing how much effort a person exerts, how much help they need, and when to hand control back to the patient. That sounds simple in a sentence and is brutally hard in hardware. ## From a monopolized market to one-fifth the price In 2017 Fourier shipped the **Fourier X1**, described in Chinese state media as the country's first force-feedback lower-limb exoskeleton rehabilitation robot. The strategic win was not just technical — it was economic. Fourier brought the price down to roughly **one-fifth of the imported equivalent**, the kind of drop that turns a luxury device into a procurement line item for ordinary hospitals. From there the company widened the catalog to **more than 40 intelligent rehabilitation products** covering upper limbs, lower limbs, balance, and physical therapy, and built a "smart rehab harbor" (智能康复港) that lets a single therapist supervise multiple devices at once. According to BrainNet (脑机网), China's AI observatory, Fourier's rehab business became the rare robotics line that pays its own bills — a cash base that later underwrote a far riskier bet. ## The humanoid detour that became the main road Gu Jie has said the humanoid (人形机器人) was a natural extension, not a pivot: if a robot can help a person stand and walk, why not one that can help them live? The timeline was longer than outsiders expected. Fourier showed a prototype only around 2022, years after it began. The products that followed: - **GR-1** (2023): a general-purpose biped with **54 degrees of freedom** and human-like gait control — among the earliest Chinese humanoids to reach mass production. - **GR-2** (2024): added dexterous hands and a rotating waist joint, improving manipulation and posture. - **Fourier N1** (2025): an open-source humanoid whose hardware drawings and code were published, a deliberate invitation to a wider builder community. There was also a softer design instinct in Fourier's more recent work. Rather than chase a hyper-realistic human look, the company has moved toward rounded contours and soft body materials in later prototypes — Gu's argument that a rehabilitation or care robot must be *touchable and approachable*, not uncanny. ## A funding base without the hype cycle Fourier is sometimes described as quieter than its flashier humanoid peers, and that fits Gu's own framing of hard-tech as a long marathon. In early 2025 the company closed a financing round of about **RMB 800 million (≈ US$111M / HK$870M)**, according to public profiles, money that supports both the rehab base and the humanoid push. Its rehabilitation products have reached medical institutions in multiple countries and regions. ## Why this story matters beyond robotics Fourier is a useful counter-narrative to the idea that Chinese robotics succeeds only through spectacle. Its first decade was built on a boring, defensible market — medical rehab — where the buyer cares about price, reliability, and clinical outcome, not demo videos. The humanoid became possible *because* the rehab business worked. ## What readers can do now - **Look at rehab robotics as the real commercial beachhead.** Before betting on humanoid (人形机器人) hype, check which companies already sell robots into hospitals today — that recurring revenue is what funds the moonshot. - **Track Fourier's open-source N1 community.** Published hardware and code mean third-party builders can now fork and extend the platform; watch GitHub and partner labs for signs of a genuine ecosystem versus a marketing gesture. - **Judge care robots by touch and trust, not looks.** If you evaluate assistive robots, prioritize soft materials, safety, and clinician adoption over how human they appear. ## Honest limitations This article draws on Chinese state-media reporting (People's Daily Shanghai, Economic Daily), encyclopedic entries, and company profiles available through late 2026; it does not include audited financial statements, which are not public for a private firm. Quantities like "40+ products," "54 degrees of freedom," and the RMB 800 million round are reported figures, not independently audited by me. I have not verified exact unit sales, hospital counts by country, or the clinical efficacy of specific Fourier devices — those would require peer-reviewed studies or procurement data I did not access. My framing favors Gu Jie's "rehab-first" logic as a sound strategy, but competitors with a demo-first approach may reach scale faster; that outcome is not settled. --- Published by NeuroAI (https://neuroai.site/) — https://neuroai.site/a/na-people-gu-jie-fourier Free to quote with attribution and a link to the original.