200+

Users a year the platform is designed to reach

50→200+

Alloys per month, year one to year six

300+

Researchers expected to be trained by year six

50%

Share of equipment time planned for outside users by year six

01

Simulate

Simulations predict what is likely to happen before any metal is poured, so the robots spend their time on experiments worth running.

02

Make & Test

Robots will melt, shape, heat-treat and test alloys around the clock.

03

Learn

AI analyzes each result and chooses what to make next, the way an experienced metallurgist would, but far faster and without delay.

Most efforts to speed up materials discovery change only the recipe: which elements go in, and how much of each. ARM-MIP will also control how the metal is made, meaning how hot it gets, how fast it cools and how it is worked. That processing history sets the metal’s internal structure, and the structure decides whether a material can survive inside a jet engine. Two pieces of metal with the same chemistry can behave nothing alike, depending on how each was made. A blacksmith quenching a blade shows the same effect.

Texas A&M to build first national self-driving laboratory for metals, open to researchers nationwide 

NSF backs Texas A&M tool that speeds materials discovery  

A faster way to find extreme-temperature alloys