This week robots with no human handler will be making their way through a field, searching for hockey pucks, blue rocks, wooden cubes and the like.
The robots won't even know what some of the objects are.
The search effort is part of the NASA's Sample Return Robot Challenge at Worcester Polytechnic Institute (WPI) in Worcester, Mass. today through Friday, where the agency hopes to find interesting technology from 18 robotics research teams.
NASA hopes the technology used at the challenge can help build improved robots for use on Mars or asteroids. The agency also hopes the new technology can help create robots that can build human habitats on other planets or moons.
A fully autonomous robot from Wunderkammer Laboratory of Canoga Park, Calif., will be competing in NASA's Sample Return Robot Challenge at Worcester Polytechnic Institute this week. (Photo: WPI)
This is the fourth Sample Return Robot Challenge has been held and the third held at WPI.
"We've been ecstatic at the progress we've been making over the past few years and can't wait to see what's going to happen this year," said Sam Ortego, NASA challenge program manager. "We need to have a system that will drive forward without falling off a cliff. We need a system that will drive forward while searching for something. All the robotic systems have to work together."
That is the tricky part for robotics.
Ortego explained that engineers have designed good sensing systems. They have good cameras. They have good mobility mechanisms. The hard part is getting it all working together.
"The concept of autonomy isn't foreign to NASA but the concept of full autonomous activity is difficult," he added. "That level of autonomy is what we need to work on."
The teams, from organizations and universities such as Rensselaer Polytechnic Institute and Oregon State University, focused on creating robots that can work to work their way around a field the size of a football field and a half, while searching for and grabbing various objects without any guidance from humans or GPS.
The easiest samples to find are worth a point each and range from a red hockey puck to an orange PVC pipe. The samples considered to be a level harder to find are worth two points a piece include a blue rock and a wooden cube.
The research teams are told what these items are so they can code them into the software, telling the robots what they should be searching for.
However, the teams have not been told about the hardest samples to find - and those are worth five points each.
So how does a robot find something if it doesn't know what it's looking for? Ortego said it needs to be programmed to recognize something that is out of place.
For instance, a robot should recognize that a pine cone would be normally found in a New England field. However, an eight-sided piece of metal would not be naturally found there so the robot should know to retrieve it.
The challenge has a total pot of $1.5 million and can be split between various teams depending on how many points they earn.
To be eligible for any prize money, a team's robot would have to retrieve at least 3 samples.
If one team earned more than 15 points, and no other team was eligible for any prize money, that team could take home the entire $1.5 million.
Sharon Gaudin covers the Internet and Web 2.0, emerging technologies, and desktop and laptop chips for Computerworld. Follow Sharon on Twitter at @sgaudin, on Google+ or subscribe to Sharon's RSS feed. Her email address is firstname.lastname@example.org.
Read more about emerging technologies in Computerworld's Emerging Technologies Topic Center.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.