AI-driven robotic lab speeds up marine materials discovery

A group of Chinese researchers has unveiled the Robotic Marine Materials Scientist Platform, an intelligent fabrication system that dramatically improves materials research and development efficiency over conventional approaches.

Traditional materials research faces two long-standing bottlenecks: valuable marine experimental data from failed trials are often lost, hindering artificial intelligence (AI) model training, while repetitive manual operations consume time and introduce errors.

To address these challenges, researchers at the Ningbo Institute of Materials Technology and Engineering of the Chinese Academy of Sciences developed the Robotic Marine Materials Scientist Platform for severe marine environments: high temperature, high pressure, high speed, high salinity, and high humidity.

The platform, covering a 400-square-meter intelligent laboratory, automates the entire workflow from design and preparation to testing and feedback. Robots act as "materials scientists" and carry out these tasks autonomously, significantly accelerating the material innovation cycle.

Working in synergy with the "MarineMat AI" assistant, the platform creates a closed-loop R&D system spanning AI-driven design, automated fabrication, characterization, and data feedback. The platform automatically records and structures all experimental parameters for model iteration, with AI prediction accuracy exceeding 90%.

Using this platform, the researchers have developed a novel multi-component cermet composite with excellent temperature and wear resistance, achieving over 30% improvement in mechanical and tribological performance over conventional ternary materials.

The screw drills made from this material have been deployed in the Chuanke-1 scientific exploration well in south China's Shenzhen at depths of 7,500 to 10,000 meters.