Tim Sprinkle

Bolts on bolts: Stanford scientist developing robotic climber

Climbing magazine
February 2004

How do you build the perfect climbing machine? Start by throwing out the inefficient two hands, two feet design you see on most modern climbers. Add some extra legs, heavy claws for grasping, a body that can pivot in ever direction, and you might be on to something. Or you could just ask Tim Bretl, a graduate student at Stanford University who has spent the last year developing a groundbreaking robotic climber that stands to revolutionize the way we think about scaling rock.

"Climbing is one of those things that humans are able to take for granted," Bretl says. "A climber just looks up at a rock and says 'there are the holds, I can go from here, to here, to here.' But when you try to synthesize that movement down into a robot, you realize that it's actually far more complex than you may have realized."

A researcher at Stanford's Aerospace Robotics Lab, Bretl has been climbing the northern California hills for several years. So when the time came to begin his doctoral work in robot design, he knew just what he wanted to do: build a machine capable of free-climbing vertical rock using the same techniques as human climbers. What he didn't know was how involved the work would become.

"Sometimes it seems like the further we get into the project, the more obstacles we find," he says, lamenting the growing pains experienced by all researchers. "There's more to it than just building a robot and programming it to climb up."

But what good is a climbing robot? Most of the Stanford drones find work as remote explorers in outer space, as Bretl's likely will someday. But he also foresees a number of more practical uses for his designs back home on Earth.

Aiding search and rescue teams - especially in remote or dangerous areas, serving as climbing porters, exploring caves and other inaccessible areas, the applications are almost limitless. For climbers, just imagine a robot capable of reaching stranded climbers with medical supplies or oxygen in an emergency; or a partner that does nothing but carry your gear all day.

"As we get further along I think we'll get a better understanding of what the robot will be capable of," Bretl says, "but until then I think we've got a lot to look forward to."

The prototype machine - a flat, octagonal body with four free-moving limbs -successfully scaled a small indoor wall last fall. The next step will be to install Bretl's new software "brain" that will allow the robot to sense and respond to the holds without outside help. Once that's been tested, practice on real rock won't be out of the question.

"We've learned a lot about how an object moves over rock," Bretl says, "how the weight sifts, how much flexibility it takes for certain moves, and a lot of that will help us when we go back to work on more traditional robot motions. Human climbers often become very aware of their movements, so we're hoping that this research will help enhance robots of all sorts."