Bringing the AlloyBot Self-Driving Laboratory to Life: Integrating Lab Automation, Computation and AI, and Developing Tungsten Alloys for Fusion Energy 100X Faster

PI: Sebastian Kube, assistant professor of Materials Science and Engineering

Co-PIs:

Joel Paulson, associate professor of Chemical and Biological Engineering

Dane Morgan, professor of Materials Science and Engineering

Co-Investigators

Adrien Couet, professor of Nuclear Engineering

Kumar Sridharan, professor of Nuclear Engineering

Oliver Schmitz, professor of Nuclear Engineering

Kirthevasan Kandasamy, assistant professor of Computer Sciences

Description: Materials that can withstand extreme temperatures, stress and radiation are urgently needed for technologies like fusion energy. Today, discovering and optimizing such alloys can take decades because experiments, simulations, and analysis are mostly conducted separately and by hand. This team’s solution, AlloyBot, is a self-driving laboratory that integrates these domains.

Robotics are used to prepare alloy samples, automated instruments measure their performance, and artificial intelligence compares results to simulations and plans the next experiments. This team will build such closed loops to test alloys over 100× faster, compressing timelines from decades to months. They will deliver proof-of-concept through a first campaign on tungsten-based alloys for fusion reactors. Their interdisciplinary team fuses automation, computation, AI and nuclear materials expertise. Seed support will allow them to implement this integration, elevate AlloyBot as one-of-a-kind capability, and build momentum as team to compete for major federal funding and forge partnerships with fusion companies.