Mifeng Technology Launches Crowdsourced Data Collection for Embodied AI Robots

Robots have not yet become common in every household, but the people who teach robots to work may already be entering ordinary homes. A new platform is turning that labor into a crowdsourced job.

On Sept. 23, Mifeng Technology launched “Mifeng Pai,” an order-based service for embodied AI data collection. Ordinary users can accept tasks, wear MEgo devices, and perform designated actions in real-world environments. The company says participants are paid according to valid collection time, making the model resemble flexible gig work.

The tasks cover homes, retail stores, warehouses, and manufacturing sites. Recruits include stay-at-home mothers, store clerks, and craftspeople, groups that can collect diverse motion and interaction data outside laboratories. Such data helps train robots to handle varied everyday objects and environments. The platform aims to scale real-scene data gathering for embodied AI training.

A Mifeng Technology executive said registration requires personal information authorization. The executive added that privacy data will be desensitized to protect users.