BOTHELL — Clark Olson was in junior high and one of the first kids on his block in Richland to get a computer more than 20 years ago.
It was a Radio Shack TRS Model 3 — a primitive tool by today’s standards — but the 14-year-old who had an interest and aptitude in math and science was hooked.
He grew up riding the wave of fast-changing technology.
Somehow, he found a way to parlay those childhood interests into a career that has literally been out of this world.
"I never thought I would be working on Mars rovers 20 years later, not even as an option," said Olson, 35, an assistant professor of computing and software systems at the University of Washington’s Bothell campus.
Yet when the Mars Exploration Rovers Spirit and Opportunity land on the Red Planet in January, Olson will have more than a passing interest. Spirit is expected to reach Mars on Jan. 4, Opportunity, Jan. 25.
For Olson, it’s not just what the rovers find but how they find it. Sure, he will be following news reports on what is discovered, but he is even more eager to learn how well they get from point to point. Helping create navigational technology for the Mars rovers has been a big part of his life.
Olson’s research is in computer vision, an area of expertise he expanded while working for NASA’s Jet Propulsion Laboratory for six years. At the UW-Bothell, his research includes visual terrain mapping for future Mars missions.
The rovers, which will land in different regions on Mars, each will carry tools to search rocks and soil for evidence about whether past environments on Mars were wet enough to support life.
A rover must be able to reach targets scientists observe from orbital or descent imagery. However, the rovers don’t have the technology to reach the distant targets with a single command. They need help. Scientists can communicate with the rovers roughly once a day because of the distance and time lag between the planets.
"Obviously, we don’t have the technology that can work as well as our eyes," he said.
Olson has worked on a new method for determining where a rover is within a map of the terrain. The method works by looking in the map to find a position that has the same geometry as the terrain that the rover can see in front of it using its cameras.
"As long as the rover has a map of the terrain, then it can tell much more accurately whether it has reached the correct location," he said. "For example, we may want the rover to place a science instrument on a particular rock. The rover can tell if it is next to the rock by determining where it is in the map."
Olson’s work is highly technical. He is the author or co-author of many articles with titles ranging from "Probabilistic Self-Localization for Mobile Robots" to "Maximum-Likelihood Image Matching," yet he can explain his work to the layman and involves two of his undergraduate students in the research.
He likes for his students at any level to be able to relate to Mars exploration. January should be a good month for that with the Spirit and Opportunity rolling across Mars.
"One of the most rewarding aspects of my research and teaching is being able to build on this excitement by bringing elements of this work into the classroom," he said.
"Students can sense this excitement, and it keeps them interested and motivated in solving complex problems."
Reporter Eric Stevick: 425-339-3446 or stevick@heraldnet.com.
Talk to us
> Give us your news tips.
> Send us a letter to the editor.
> More Herald contact information.
