If you place a rigid sphere on a similarly rigid inclined surface, gravity will cause it to roll down said surface. But what happens if the surface, or plane, is completely vertical? Researchers had previously assumed that, without an initial push, the sphere would simply drop straight to the ground without rolling. New research, however, has just redefined this belief—as well as long-held assumptions in the field of physics.

University of Waterloo researchers have revealed the exact scenario necessary to make a sphere roll down a vertical plane without physical intervention. While this niche observation might seem detached from everyday life, it could have useful applications for exploring hard-to-reach areas such as pipes, caves, and even space.

“When we first saw it happening, we were frankly in disbelief,” Sushanta Mitra, executive director of the Waterloo Institute for Nanotechnology, said in a university statement. The researchers describe their discovery as a challenge to “our basic understanding of physics.” They “double-checked everything because it seemed to defy common sense. There was excitement in the lab when we confirmed it wasn’t a fluke and that this was real vertical rolling.”

Mitra and his colleagues unexpectedly caught the vertical rolling with high-speed cameras, and explained their discovery in a study published in April in the journal Soft Matter.

In their experiment, the vertical rolling depended on a precise balance of softness—scientifically defined as elasticity—between a small sphere and a vertical cellphone-sized surface. When the spheres were too solid, they simply fell directly to the ground. On the other hand, when they were too soft, they either slid down without rolling, or stuck to the plane. But a sphere about as soft as a gummy bear spontaneously rolled down a vertical surface equivalent to a spongy mouse pad at a speed of about 0.039 inches (one millimeter) every two seconds, as described in the statement.

“The key is that as it rolls, the sphere slightly changes shape at the contact point,” Mitra explained. “The front edge acts as a closing zipper, while the back edge acts like opening it. This asymmetry creates just enough torque, or grip, to maintain rolling without either sticking or completely falling off.”

The team’s findings could have practical implications for the creation of soft robots that can scale vertical walls to explore or monitor inaccessible infrastructure and natural environments both on and off Earth. “This opens up a whole new way of thinking about movement on vertical surfaces,” Mitra continued. “Currently, robots and vehicles are limited to horizontal or slightly inclined surfaces. This discovery could change that.”