How the Complicated Physics Inside Your Coffee Mug Could Lead to Better Robots

The natural human gait walking along a straight path (x) affects the trajectory of the coffee (z) in the cup based on the velocity of the gait.

- By Courtney Linder

DESPITE WHAT THAT LONG-FORGOTTEN stain on your white shirt might have you believe, humans are pretty good at walking with a cup of coffee and avoiding spills, even if our success rate isn't quite 100 percent. Every time you manage to get your cup of joe from one point to another spill-free, you're intuitively completing a little-understood feat of physics: manipulating a complex object such as liquid.

That's according to a group of researchers at Arizona State University (ASU) who have been modeling the coffee-carrying phenomenon in an attempt to imbue robots with the same finesse. In a world of increasing automation, machines are expected to perform more dexterous motions, explains Brent Wallace, a Ph.D. student at ASU's School of Electrical, Computer, and Energy Engineering who was involved in the work.

“But even for simple tasks, like carrying a cup of water or a cup of coffee, the robot struggles. Every day, you and I make a cup of coffee, and 99 out of 100 days, we don't spill it on ourselves, Wallace says. “So how do we get leverage on tackling those kinds of problems? Well, let's study how humans behave in those situations. Building on prior work at Northeastern University, which found that humans have two main approaches in manipulating a complex object like a fluid, the ASU team simulated those responses, lasering in on the transition phase between the two to understand why humans exhibit a binary response-and to see how robots could learn to do the same in the future. The findings were published in the journal Physical Review Applied in late 2021.