Eidon AI, a robot data company, has announced its closure, citing the inability of the revenue generated from embodied AI training data to cover operational expenses. As a result, the company has decided to open-source its technology and the data it has amassed over more than two years. After shifting its focus to the robot data business, Eidon AI encountered numerous challenges, including difficulties in data collection and operations, hardware iteration, production capacity supply, and bargaining power. These hurdles ultimately led to the company's inability to sustain its operations.

Currently, while capital continues to flood into the overseas embodied AI data sector, there are signs of oversupply and falling prices. Leading downstream robotics companies are increasingly establishing their own data channels, thereby constraining the survival space for independent data providers. In the domestic market, a capital boom is still underway, with independent data service providers enjoying high levels of financing and pricing. However, they too will face downward pricing pressure in the future.

Typically, third-party data collectors have a development lifespan of only 1-3 years. Relying solely on selling data is not a viable long-term strategy. To achieve sustainable growth, they need to diversify into areas such as simulation, evaluation, or model training.