Mifeng Technology Launches Mifengpai Physical AI Data Crowdsourcing Platform and Robot Trainer Program
Mifeng Technology launched Mifengpai in Shanghai, calling it the world's first all-category physical AI data crowdsourcing platform. It also announced a robot trainer career plan, a 100 million yuan subsidy package and a data governance initiative.
Mifeng Technology Chairman and CEO Yao Maoqing said at the event that the obstacle to robots entering factories, stores and homes is not algorithms or hardware but a lack of massive, real human action data. According to the company, Mifengpai converts ordinary actions such as folding clothes, brewing tea and stocking shelves into high-value data for robot training, turning the slogan everyone can be a teacher into a practical system.
The platform has two layers. Its infrastructure includes MEgo-series collection devices, the Mifengpai app and the MEgo Engine data post-processing platform. Its upper layer is a scene network and a collector network, based on the idea that wherever real work happens is a data origin and that ordinary people can take orders, collect data and settle payments at any time. MEgo View is described as an industry-first full-sensory, multi-view collection device, with head-mounted multi-camera coverage of more than 300 degrees. MEgo Gripper offers millimeter-level reconstruction accuracy and optional three-dimensional tactile sensing.
Users download the Mifengpai app, complete real-name registration and apply for a device. They then follow four steps, taking orders, collecting data, having it accepted and settling payment. Withdrawals are supported through WeChat, Alipay and UnionPay and arrive within two minutes, with taxes paid automatically by the platform, according to the company. Mifengpai currently covers 22 major categories, more than 5,000 task types and over 50,000 real scenes. During a beta test from Aug. 15 to Sept. 15, nearly 20,000 users registered, and the top user earned 5,111 yuan in a single month.
At the launch, 16 beta users received what the company called the world's first robot trainer certifications. Mifeng Technology also released a robot trainer career plan for the broader public. The Ministry of Human Resources and Social Security and other departments had previously added embodied intelligent robot application technician to the list of new occupations. The company said the launch aligns with the national industrial direction.
Mifengpai Business Vice President Zhang Zhifu announced a 100 million yuan subsidy plan. It includes 50 million yuan in task subsidies, 20 million yuan in equipment subsidies, 30 million yuan to build an offline service network covering 40 cities, and 2 million yuan for dual insurance covering both people and devices. The company also introduced an M1 to M5 five-level growth system and set a two-year target of 1 million daily active users and 1 million robot trainers.
Industry ecosystem and governance efforts were launched at the same time. Mifeng Technology and the China Academy of Information and Communications Technology began compiling the Embodied Intelligence Crowdsourced Data Development Report, with co-builders including the National Robot Testing and Evaluation Center, Guizhou Big Data Group and China Telecom Data Development Center. A Scene Data Alliance was established on site, with 12 first members including Longcheer Technology, Huayi Hotel and Deji Hospital. Mifeng Technology technical lead Liu Li read an action initiative on embodied intelligence data security and protection, setting rules for collection standards, full-chain security and industry coordination.
At the launch, the company said global high-quality real physical interaction data amounts to only about 100,000 to 1 million hours, while robot commercialization requires tens of millions to hundreds of millions of hours, leaving a gap of more than 99 percent. Crowdsourced collection is becoming a new path to expand data supply. Yao said the experience of billions of hands will eventually converge into robots' skills. Mifengpai is now operating in 40 Chinese cities, and the company said it will work with industries to complete the physical AI scene data map and accelerate the arrival of the AGI era.