A community perspective piece to which I contributed has been published in Digital Discovery.

Self-driving labs (SDLs) combine artificial intelligence, automation, and advanced computing to accelerate scientific discovery. This paper, authored by a broad cross-disciplinary group from academia, government, and industry, takes stock of where the field stands and looks honestly at what remains to be solved.

Rather than celebrating recent successes, the perspective focuses on barriers, blind spots, and open challenges — from limitations in current AI and automation strategies to data infrastructure, reproducibility, and the social dimensions of autonomous discovery. The goal is a clear-eyed view of what the community needs to address to move from promising demonstrations toward reliable, general-purpose self-driving laboratories.

The paper grew out of discussions at the inaugural Accelerate Conference and reflects the collective voice of a community that is enthusiastic but appropriately cautious about the road ahead.