RICE AI functions as a platform dedicated to crowdsourced robotics data collection. It allows teleoperators to manage robots remotely, capturing information from visual feeds, joint articulations, and force sensors. Operators utilize tools such as webcams equipped with skeleton detection, joysticks, VR controllers, or specialized teleoperation rigs, receiving token rewards that correlate with the dexterity of the device used.

Both individuals and enterprises can supply robots—including humanoid units, mobile bimanual systems, quadrupeds, and AI toys—to specific data collection hubs in exchange for tokens. Devices in higher demand yield greater rewards. These hubs are situated in either actual commercial environments or laboratory settings.

The aggregated data is sold to corporate clients and research institutions. Additionally, RICE AI employs this data to train robotics foundation models, which are accessible through a monthly subscription service. Users who pay with the native token enjoy a discount. Revenue generated from data sales facilitates token burns, thereby decreasing the total supply, while the token also acts as a governance tool for the ecosystem.

To ensure data integrity, an AI evaluation model reviews contributions from teleoperators, aided by human labelers who rank samples according to task completion efficiency. Furthermore, an AI system predicts similarity between data episodes, and a clustering-based embedding database removes redundant entries to uphold quality standards.

The total supply of RICE is capped at 1 billion. However, when the Token Generation Event occurred, the circulating amount stood at approximately 187 million.