A company born from the founder’s own reality
On 16 October, ZhuGe Robotics of Shandong closed a Pre-A round with Pudong Venture Capital, Bawei Tong Spatial Intelligence and Shijiazhuang Quchuang participating. At a moment when most embodied-intelligence projects are still stuck at lab demos, a blind-assistant robot aimed at the visually impaired drew both industrial and state-backed capital, and that shift reveals a structural change in the track: accessibility scenarios are moving from a public-welfare concept to a commercial bet investors will fund.
The company’s DNA is stamped by founder Zhu Qingyi’s own experience. A high fever at fifteen dropped his vision to 0.01. He did not let it trap him. He entered Changchun University’s Special Education College, taught himself programming, and in 2001 built China’s first accessibility website. Over the next two decades he stayed deep in information accessibility, helped draft several national and international standards, and became a specially appointed expert on accessibility technology for the China Disabled Persons’ Federation.
Why capital is drawn in
Because he is himself part of the visually impaired community, Zhu understands the small, lethal obstacles of a trip outside better than any outside team, and that pushed the company from software accessibility toward hardware robotics.
Pudong Venture Capital, a state-backed platform, invests across hard tech and AI, and its logic here was not flashy specs but the closed loop of technology, data and social value an accessibility scene can form. Bawei Tong Spatial Intelligence brought more than money: its spatial-intelligence algorithms and compute base let the two sides link city-space perception with the robot’s data pipeline for joint R and D. Industrial capital adds algorithms, compute and sector resources, not just a balance-sheet entry. The hardest problem for an embodied-intelligence startup was never a prototype but a scene and a data source to keep iterating, and ZhuGe stepped on exactly that.
What makes the guide robot hard
Many reduce a blind-assistant robot to an enlarged navigation device, but real city pavements are far more complex than phone navigation. Dropped obstacles, temporary piles, crossing pedestrians and riders are all live risks a traditional navigator cannot see. The AI LOOK robot’s key difference is escaping the pre-scan-and-remote-control pattern. Many service robots need staff to map an area first, and any change breaks the system, which cannot fit a thousand shifting public streets.
Relying on multi-sensor fusion and a pedestrian-corridor visual model, the robot perceives, decides and guides in unfamiliar environments on its own, judging passable areas and predicting hazards instead of merely capturing images. The team balanced compute heat and outdoor waterproofing through repeated real-machine testing. The real accumulation of real users’ data in real streets is the actual barrier, and every outing feeds the model.
A possible breakthrough path
The signal this round sends goes beyond one company’s funding. For a long stretch the domestic industry chased general humanoids and flashy demos while cost and missing scenarios blocked commercialisation. Accessibility offers another route: precise, rigid demand, clear pain, public and campus and ageing-care markets that open commercial imagination. The track still faces cost, standards and scale hurdles, but after the round ZhuGe plans to keep iterating its model, harden its scene-data moat and widen its reach across disability and ageing care.
Editor’s note: This is an adapted translation of the original OFweek Robotics report. It has been trimmed and restructured for readability for an international business audience. The full original (in Chinese) is at https://robot.ofweek.com/2026-10/ART-8321203-8420-30707450.html.