Embodied Intelligence Observer

Jianzhi Robotics简智机器人

LatestAngel (2026, led by Shunwei Capital)

Undisclosed (to be confirmed)

Jianzhi Robotics (full registered name to be confirmed)

An embodied AI data-infrastructure startup focused on human data collection, producing robot data via 'wearable collection devices + a crowdsourcing model'; completed 3 rounds totaling over RMB 200M within 4 months of founding

Embodied ModelsChina

数据截至2026-08-19

FoundedJul 2025 (per original notes; completed 3 funding rounds within 4 months of founding, to be confirmed)
HQTo be confirmed (per original notes)
FounderZhu Yanming
Products5 items
Funding Rounds3 rounds
Events2 items

About

Jianzhi Robotics was founded in July 2025 (to be confirmed) and is a representative company in the embodied AI data infrastructure track, focused on human data collection and producing robot training data via 'wearable collection devices + a crowdsourcing model'. Its core product, the Gen DAS wearable device, weighs about 470 g with trajectory precision <1 cm, multi-device synchronization latency <1 ms, and over 30,000 detection points on its tactile sensors; it is paired with the Gen Controller skill-collection terminal, the Gen Matrix data platform and the Gen ADP data pipeline, forming a '1+x' product system. The crowdsourced data network has accumulated over 1 million hours of human data covering 500+ skills, mainly serving embodied AI model trainers. Within 4 months of founding the company completed 3 rounds totaling over RMB 200M: a Seed round led by Baidu Ventures (with Momenta, 九识智能 and 星海图 following), a Seed+ round led by RoboSense, and an Angel round led by Shunwei Capital — deeply tied to industry-chain players such as 星海图, RoboSense and Momenta. The team has about 140 members (per original notes); specific orders and customers are yet to be disclosed.

Technical Approach

Uses a human-data crowdsourcing model to collect human manipulation data at scale, with over 1 million cumulative hours covering 500+ skills; achieves device-level data standardization with high-precision wearables (<1 cm trajectory precision, <1 ms multi-device sync) to ensure data quality and multimodal alignment; industrializes data production into pipelines, solving embodied AI's 'data hunger' problem.

Commercialization(3)

  • The crowdsourced data network has accumulated over 1 million hours of human data across 500+ skills (per original notes)
  • Main customers are embodied AI model trainers; the business model is data collection and pipeline services (to be confirmed)
  • Specific orders and customer information to be disclosed

Position & Strengths(3)

  • A representative embodied AI data-infrastructure company, deeply tied to industry-chain players such as 星海图, RoboSense and Momenta (via investment + cooperation)
  • Differentiation: mass-produces high-quality human data with a 'wearable device + crowdsourcing' model, with outstanding cost and scale advantages
  • Backed by Baidu Ventures across three consecutive rounds (Seed, Seed+, Angel), plus support from Shunwei Capital and 初心资本 — strong capital recognition

Links1