The lab that learns · M. Abolhasani
原文开头
Integrating high-throughput experiments, robotics, and artificial intelligence can accelerate scientific discoveries Milad Abolhasani The traditional approach to discovering materials or molecules involves an iterative process in which one combination is created and tested until desired properties are reached. Developing a useful material or molecule can take years or decades of synthesis, characterization, failure analysis, and redesign. The challenge is that composition alone rarely determines performance. In traditional discovery, human judgment helps determine which variables are explored, which results are prioritized, and which candidates are tested next. Self-driving laboratories offer a different model for discovery. …
摘自《科学》(Science)第393卷 第6813期 · 2026年8月20日,M. Abolhasani。仅引用开头一小段供了解文章,版权归原刊所有,全文请阅读原刊。