Skip to main content


[an error occurred while processing this directive]
[an error occurred while processing this directive]
[an error occurred while processing this directive]
[an error occurred while processing this directive]

Building Better Robots

"The best thing about robotics is that I can play with toys and still try to contribute something meaningful to society," says Nelson Rosa Jr. '05, now a B.E. candidate at the Thayer School of Engineering.

Rosa started working in Dartmouth's Robotics Laboratory during the fall term of his senior year after taking an electronics course. He and research partner Joshua Pyke, also an '05 pursuing a B.E. at Thayer, teamed up to tackle a seemingly basic problem: how to free a remote-controlled robotic vehicle stuck in the sand.

Joshua Pyke '05, Devin Balkcom, and Nelson Rosa Jr. '05
Joshua Pyke '05 (left), Assistant Professor of Computer Science Devin Balkcom (center), and Nelson Rosa Jr. '05 in the Sudikoff Robotics Laboratory.

The challenge, as presented by Robotics Laboratory Director and Assistant Professor of Computer Science Devin Balkcom, was to make a simple robot do complicated things.

"When wheeled vehicles get stuck," Balkcom says, "the first order of business is to not make the situation worse."

For instance, if a robot got stuck on the loose, soft surface of Mars, the people at the controls back on Earth would want to make sure they didn't compound the problem in their attempt to free the vehicle. (This actually happened to the Mars Rover.) The same goes for a remote-controlled truck stuck in the sand while perhaps on a rescue mission in the Mojave Desert. The goal is to regain mobility, not dig a deeper hole.

"My biggest success was getting the robot to dig itself out of a sand trap," says Rosa. "I was really excited to see the wheels of the robot, after being buried in sand, rise out and get going again."

Rosa and Pyke, with guidance from Balkcom, employed two methods to free their machine: oscillating and digging. Now that they've demonstrated it can be done, the next step is to vary the sand, make it finer, coarser, or deeper, and add a few sensors that could aid in the escape. The trick, though, is to develop a robot that isn't burdened by too much extra equipment.

"We could just add a huge mechanical arm to the robot," says Balkcom, "but that adds to the cost and the weight of the machine. We are minimalists."

Undergraduates, graduate students, and post-doctoral fellows all work at the Robotics Laboratory in Sudikoff, home to the Department of Computer Science. Other projects currently under way involve developing robots that can climb with just a few sensors, building a robot that can fold origami (an extension of Balkcom's doctoral thesis), and programming a robot to carefully dismantle structures without a collapse.

Rosa sums up his work with robots as a marriage of his interest in machines that have a practical use and his aptitude for electronics and computer science: "Robotics just happens to be that happy intersection where I can apply both simultaneously to create a smart piece of machinery."


Questions or comments about this article? We welcome your feedback.

Last Updated: 5/30/08