Skip to main content
Argonne National Laboratory

Developing a self-driving laboratory prototype

Argonne is integrating instruments and robotics into a prototype AI-driven, autonomous laboratory that will allow rapid exploration of innovative ideas and scientific breakthroughs.

Self-driving laboratories (SDLs) can perform experiments without human intervention, under computer control. Experiences in the manufacturing industry show that even extremely complex processes can be automated reliably and rapidly. However, while the goal in manufacturing is typically to repeat certain steps over and over, success in scientific research involves combining different elements in new ways.

Argonne is working to develop its own prototype SDL — a vibrant software and robotics hub where scientists from across the laboratory can collaborate, train the next-generation workforce, drive innovation, and foster an open-source environment. Establishing this laboratory will enable students, scientists, and engineers to rapidly experiment with new ideas in an environment where production-level operations are not required.

The SDL prototype will focus on developing robotics software at all levels — from instrument drivers to high-level user programming APIs (application programming interfaces — which allow two applications to talk to each other). The goal is to integrate instruments and robotics into a prototype AI-driven, autonomous laboratory.

Such a prototype is essential because significant software engineering is required to configure new systems that contain fabricated components and to develop protocols. Early stages of software development typically involve considerable testing and debugging. These steps are time consuming and often cause things to break, which would have  significant negative effects on user experiments in a production environment.

The prospective benefits of the Argonne SDL prototyping laboratory extend beyond software and robotic development to include development of new talent and fostering of new collaborations. Shifting to open-source rather than vendor-specific software will greatly reduce the time and expense of instrument and robotic integrations, enabling rapid exploration of innovative ideas and breakthroughs in both our SDLs and in the science that our SDLs support.