Facebook
Britain's News Portal
Around The Clock
BREAKING
Loading latest headlines…

Physical AI faces 'data crisis' despite investment boom and IPO volatility

The physical AI sector is experiencing significant investment, but a lack of high-quality training data is hindering the development of commercially reliable robots. This challenge is highlighted by the recent halving of Unitree's market value.

  • Unitree, a leading Chinese robot maker, saw its market value nearly halve after its IPO, with analysts citing the robots' lack of value-creating know-how.
  • Developers at the Actuate conference identified a 'robotics data crisis' due to insufficient high-quality training data for AI models.
  • Some companies are focusing on specific tasks to get robots into the field, while others are developing simulation tools and data management solutions.

The physical AI sector, which applies large language model tools to robotics, is attracting substantial venture investment. However, the industry is grappling with a significant challenge: a shortage of high-quality training data for AI models, which is impacting the commercial performance of robots.

This issue was underscored by the recent performance of Unitree, a prominent Chinese robot manufacturer. After its IPO, which initially valued the company at $66 billion, its market value nearly halved. Analysts attribute this decline to the robots' current inability to perform value-creating work effectively.

At the recent Actuate conference, a gathering for developers creating AI brains for robots, the 'robotics data crisis' was a key concern. Attendees discussed the need for more diverse datasets, varied training regimes, and improved reinforcement learning scenarios to advance physical AI beyond its current 'GPT 2 era,' a reference to an earlier OpenAI model.

Companies are exploring different strategies to address these challenges. Some, like Gritt, Agility, and Bedrock, are focusing on specific tasks to deploy robots in real-world settings, such as building solar farms or operating excavators. Others, including Foxglove, are developing tools to help manage and visualize the dense visual and lidar data required for physical AI models.

Why this matters: The development of physical AI could lead to robots capable of performing various tasks, but the current 'data crisis' indicates significant hurdles remain before widespread commercial application.

Related Articles

Get the news that matters.

Join thousands of readers getting the best of British news straight to their inbox.