AnswersPublished19 sources
Maniformer’s Mifengpai Turns Everyday Tasks Into Robot-Training Data
Mifengpai lets participants record tasks such as folding clothes and making tea with MEgo devices, earning payment for data that passes review. The platform combines wearable capture hardware, an app and MEgo Engine processing to turn real world demonstrations into structured data.
Published byEdited with GPT-6 LunaImages generated with GPT Image 2
Research answer
Key takeaways
- Mifengpai lets participants record tasks such as folding clothes and making tea with MEgo devices, earning payment for data that passes review.
- The platform combines wearable capture hardware, an app and MEgo Engine processing to turn real world demonstrations into structured data.
- Its reported beta sign ups and submissions show early activity, but do not establish data quality or improved robot performance.
AI Prompt
openai.comCreate a landscape editorial hero image for this Studio Global article: How does Shanghai-based physical AI training-data company Maniformer’s crowdsourcing platform, launched in Shanghai on September 23, 2026, u. Article summary: Maniformer’s platform, Mifengpai (觅蜂派), aims to turn ordinary people’s real-world work into paid robot-training demonstrations, addressing CEO Yao Maoqing’s argument that the missing ingredient is abundant human-motion d. Topic tags: general, general web, user generated. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fa
Maniformer launched Mifengpai in Shanghai on September 23, 2026, with a simple proposition: people record everyday actions, and those recordings may help train robots. CEO Yao Maoqing has argued that robots need more real human-motion data to move into factories, shops and homes. Mifengpai is the company’s attempt to gather that data beyond the robotics lab. 30
From everyday work to robot data
Participants use the app to choose a task, record it with MEgo equipment, and submit the data for review. Payment is based on the amount of recording time approved as usable—not simply the time spent doing the task. 19
25
41
Launch coverage describes tasks such as folding clothes, making tea and arranging goods. The idea is to capture how people perform actions in real settings, where surroundings and interactions can differ from a controlled demonstration. 30
The equipment and processing behind the platform
Mifengpai combines MEgo capture devices with a mobile app and MEgo Engine. Launch reporting describes MEgo View as a head-worn device whose cameras cover more than 300 degrees, and says MEgo Gripper can reconstruct manipulation movement at millimeter scale. 2
30
The system is designed to gather more than ordinary video: reporting describes recording visual information alongside movement and interaction data. MEgo Engine then processes and labels submissions and assesses their quality, preparing data for potential model training or evaluation. 25
36
45
In other words, the platform connects people doing real tasks with equipment and software intended to turn their demonstrations into structured data. That is the proposed answer to the data shortage Yao has identified—not, by itself, proof that the shortage has been solved. 30
What the early numbers do—and don’t—show
Launch coverage reported 20,000 registered users and 13,000 task submissions during a one-month beta. These figures suggest early participation, but they do not show how many submissions passed review, how much usable data was produced or whether the recordings improved robot performance. 19
41
Other reports list 22 categories, more than 5,000 task types and over 50,000 real-world settings. These are reported platform figures, not independent assessments of data quality or robot capability. 7
14
The open question
Mifengpai offers a way to collect human demonstrations at scale and pay contributors for recordings that pass review. The available launch reporting explains how the platform is intended to work and describes its early activity; it does not independently verify the quality of accepted submissions, long-term participant earnings or gains in robot capability. For now, the platform is a proposed approach to a data bottleneck—not evidence that the bottleneck has been overcome.
Studio Global AI
Continue your research
This page includes a source-backed answer you can continue inside Studio Global.
People also ask
What is the short answer to "Maniformer’s Mifengpai Turns Everyday Tasks Into Robot-Training Data"?
Mifengpai lets participants record tasks such as folding clothes and making tea with MEgo devices, earning payment for data that passes review.
What are the key points to validate first?
Mifengpai lets participants record tasks such as folding clothes and making tea with MEgo devices, earning payment for data that passes review. The platform combines wearable capture hardware, an app and MEgo Engine processing to turn real world demonstrations into structured data.
What should I do next in practice?
Its reported beta sign ups and submissions show early activity, but do not establish data quality or improved robot performance.