The experimental sciences have long relied on labor-intensive, manual workflows that limit scalability and slow the pace of research and discovery. Despite advances in instrumentation and data analysis, the act of performing physical experiments remains time-consuming and resource- intensive, ultimately limiting the pace of exploration and innovation. The emerging field of self- driving laboratories (SDLs) and autonomous labs seeks to overcome these challenges by integrating robotics, automated experimentation, and artificial intelligence (AI) to design and execute experiments with minimal human intervention. Through this integration, SDLs are redefining how research is conducted across scientific domains, enabling a more adaptive, data- driven, and efficient approach to scientific discovery.
In this talk, we first outline the foundational principles of SDLs and discuss the critical transition from automated experimentation to autonomous experimentation. Next, through a series of case studies, we illustrate how increasing levels of autonomy transform experimental workflows and enable systems that not only execute tasks but also make data-driven decisions to guide discovery. We then examine emerging challenges in autonomous laboratories, particularly the limited flexibility and adaptability of systems that are often designed for narrow or limited experimental workflows. As SDLs expand in scale and complexity, they become increasingly vulnerable to unanticipated faults, underscoring the need for resilient architectures that maintain performance under changing or complex conditions. Finally, we highlight how integrating approaches such as tele-robotic operation, large language models (LLMs), and agentic AI frameworks may enhance adaptability and fault tolerance in these systems. Together, these advances pave the way for the next generation of SDLs, where intelligence and reasoning span both the system and instrument levels of autonomous labs.
Aldair Gongora is a staff scientist at Lawrence Livermore National Laboratory (LLNL) and also serves as a deputy director for LLNL’s Data Science Institute’s Consulting Service. His research work centers on accelerating scientific discovery by developing next-generation research platforms and capabilities, such as self-driving labs and autonomous robots for science. His research work enables the adoption of modern tools to tackle complex challenges across materials science, life sciences, and advanced manufacturing. Aldair earned a B.S. in Mechanical Engineering from Rockhurst University and a Master's and Ph.D. in Mechanical Engineering from Boston University, where he conducted research on self-driving labs and autonomous experimentation for mechanical design.