Centennial - Blog Posts

Sample Return Robot Challenge

It’s been a long, technical journey for the seven teams competing this week in Level 2 of our Sample Return Robot Challenge. Over the past five years, more than 50 teams have attempted the $1.5 million competition, which is looking to develop autonomous capabilities in robotics. Basically, we want robots that can think and act on their own, so they can travel to far off places – like Mars – and we can rely on them to work on their own when a time delay or unknown conditions could be factors.

This challenge has two levels, both requiring robots to navigate without human control and Earth-based tools (like GPS or magnetic compassing). The robot has to find samples, pick them up and deliver them to home base. Each of the final seven teams succeeded at Level 1, where they had to find one sample, during previous competition years. Now, they have a shot at the much more difficult Level 2, where they have a two-hour window to locate up to 10 samples of varying point values, but they don’t know where to look or what exactly they’re looking for.

Get to know the final seven, and be sure to cheer them on as we live-stream the competition all day Sept. 4 and 5.

West Virginia University Mountaineers Hailing from: Morgantown, West Virginia # of Team Members:  12

Behind the Name: In West Virginia, we call ourselves mountaineers. We like to explore unknown places and be inspired by nature.

Motivation: To challenge ourselves. Through this venture, we are also hoping to create research and career opportunities for everyone on the team.

Strategy: Keeping things simple. Through participating in SRR challenge during the last three years, we have gone a long way in streamlining our system.

Obstacles: One of the biggest challenges was finding and nurturing the talent of individual team members and coordinating the team in making real progress on time.

Prize Plans: We donated 50 percent of our 2015 Level 2 prize money to create an undergraduate “Robotics Achievement Fellowship” at WVU. The rest of the funding was allocated to support team member professional development, such as traveling to conferences. A similar model will be used if we win in 2016.

Extra Credit:  We did an Easter egg hunt with our robot, Cataglyphis (named after a desert ant with extraordinary navigation capabilities), last year.

Survey Hailing from: Los Angeles, California # of Team Members: Jascha Little

Behind the Name: It's short, simple, and what the robot spends a lot of its time doing.

Team History: We work together, and we all thought the challenge sounded like an excellent way to solve the problem of what to do with all our free time.

Motivation: We are all engineers and software developers that already work on robotics projects. Reading too much sci-fi when we were kids probably got us to this point.

Strategy: We are trying to solve the search-and-return problem primarily with computer vision. This is mostly to reduce cost. Our budget can't handle high quality IMUs or LIDAR.

Prize Plans: Probably build more robots.

Extra Credit: Favorite pop culture robot is Bender (Futurama). Alcoholic robots are the best.

Alabama Astrobotics (The University of Alabama) Hailing from: Tuscaloosa, Alabama # of Team Members: 33

Behind the Name: “Alabama Astrobotics” was chosen to reflect our school affiliation and our mission to design robotics for various space applications.

Team History: Alabama Astrobotics has been involved with other NASA robotics competitions in the past.  So, the team is accustomed to the competition environment.  

Motivation: We are pleased to have advanced to Level 2 in our first year in the competition (the first team to do so), but we are also not satisfied with just advancing.  Our goal is to try to solve Level 2.

Strategy: Our strategy is similar to that used in Level 1.  Our Level 1 approach was chosen so that it would translate to Level 2 as well, thus requiring fewer customizations from Level 1 to Level 2.

Obstacles: As a university team, the biggest challenge was not having all our team members available to work on the robot during the time since Level 1 completed in June. Most of my team members have either graduated or have summer internships, which took them away from campus after Level 1.  Thus, we didn’t have the manpower to address the additional Level 2 technical challenges.

Prize Plans: Any prize money would be donated to the University of Alabama College of Engineering.

Extra Credit: Alabama Astrobotics also competes in the annual NASA Robotic Mining Competition held at the Kennedy Space Center each May.  We have been fortunate enough to win that competition three times in its seven year history, and we are the only team to win it more than once.

MAXed-Out Hailing From: Santa Clara, California # of Team Members: 4

Behind the Name: Several reasons: Team leader is Greg Maxwell, and his school nick name was Max. Our robot’s name is Max, which is one of the most common name for a dog, and it is a retriever. Our efforts on this has been too the max…. i.e. MAXed-Out. Our technology requirements have been pushed to their limits - Maxed-Out.

Team History: Greg Maxwell started a Meet-up “Silicon-Valley Robot Operating System” SV-ROS that was to help teach hobbyists how to use ROS on their robots. We needed a project to help implement and make real what we were teaching. This is the third contest we have participated in.

Motivation: There is still such a long way to go to make robots practical. Every little bit we can contribute makes them a little bit better and smarter. Strategy: Level 1 was a test, as a minimum viable product to prove the tech worked. For Level 2, we had to test and add obstacle avoidance to be able to cover the larger area with trees and slopes, plus add internal guidance to allow for Max to be out of the home base camera tracking system.

Obstacles: Lack of a cost effective robot platform that met all the requirements; we had to build our own. Also time and money. The two months (between Level 1 and 2) went really fast, and we had to abandon lots of cool ideas and focus on the basics.

Prize Plans: Not sure, but pay off the credit cards comes to mind. We might open-source the platform since it works pretty well. Or we will see if it works as expected. We may also take a break / vacation away from robots for a while.

Extra Credit: My nephew, Max Hieges, did our logo, based on the 1960-era Rat Fink sticker.

Mind & Iron Hailing From: Seattle, Washington # of Team Members: 5

Behind the Name: It was the original title for Isaac Asimov’s “I Robot,” and we thought it was a good combination of what a robot actually is – mechanical and brains.

Team History: Three of us were WPI undergrads and met at school; two of us did our master’s degrees at the University of Washington, where we met another member, and then another of us brought on a family member.

Motivation: We saw that there was an opportunity to compete in a challenge that seemed like there was a reasonable solution that we could tackle with a limited budget. We saw three years of competition and thought that we had some better ideas and a pretty good shot at it. Strategy: The samples and the terrain are much more complex in Level 2, and we have to be more careful about our navigation. We are using the same tools, just expanding their capability and scope.

Obstacles: The team being spread over three different time zones has been the biggest challenge. We are all doing this in our free time after work. The internet has been really handy to get things done.

Prize Plans: Probably invest in more robot stuff! And look for other cool projects we can work on, whether it’s another NASA challenge or other projects.

Extra Credit: We are hoping to collaborate with NASA on the professional side with surgical robots to exoskeletons. Challenge-related, our robot is mostly made of plywood – it is a composite fiber material that works well for fast development using cheap materials.

Sirius Hailing From: South Hadley, Massachusetts # of Team Members: 4

Team History: We are a family. Our kids are both robot builders who work for Boston Dynamics, and they have a lot of robot expertise. Both of our kids are robotics engineers, and my wife is intrinsically brilliant, so the combination of that makes for a good team.

Motivation: Because it’s a really hard challenge. It’s one thing to drive a robot with a remote control; it’s another to do the whole thing autonomously. If you make a single change in a robot, it could throw everything off. You have to think through every step for the robot. On a basic level, to learn more about robotics and to win the prize. Strategy: Very similar to Level 1. We approached Level 1 knowing Level 2 was there, so our strategy was no different.

Obstacles: It is very difficult to do object recognition under unpredictable conditions – sun, clouds, weather, sample location. The biggest challenge was trying to recognize known and unknown objects under such a wide variety of environmental possibilities. And the terrain is very different – you don’t know what you’re going to find out there.

Prize Plans: We haven’t really thought about it, but we will give some away, and we’ll invest the rest in our robotics company.

Extra Credit: The first robot we had was called Robo-Dad. Dan was training to be an astronaut in the 1990s, so we built a toy remote-controlled truck that Dan - in Texas - could control via the internet in the house. Robo-Dad had a camera that Dan could see the house with. It had two-way communication; it was a little before it’s time – the internet was very slow.

Team AL Hailing From: Ontario, Canada # of Team Members: 1

Team History: I was looking for competitions that were open, and my dad had followed the Centennial Challenges for a while, so he alerted me to this one. I was already doing rover projects, and it was appropriate and awesome and interesting. I felt like I could do it as a team of one.

Motivation: Difficult challenges. I’m definitely inspired seeing really cool robots that other people are building. New emerging tech really motives me to create new things.

Strategy: I showed up with another robot to Level 2. I built three, but ran with only two. It did make it more complicated, but the strategy was to send them to different areas and have them be able to communicate with each other. Everything physically was the same from Level 1.  The idea is that they would all go out with different missions and I would maximize field coverage.

Obstacles: Time. More time would always be nice. Being able to make something like this happen under a timeline is really difficult. I feel like I accomplished a lot for a year. Also, manpower – being a team of 1, I have to do all of the paperwork and other related stuff, but also carry the hardware and do the programming. You have to multitask a lot.

Prize Plans: I’d like to start a robotics company, and be able to expand some of the things I’ve been working on associated with technology and maker education.

Extra Credit: My story is not linear. A lot of people are surprised to hear that my background is in molecular biology and  research. I once lived in a tent in Madagascar for a few months to do a biodiversity study, and I have multiple publications from that side of my life. I am in a whole different place now.

The competition is one of many run by our Centennial Challenges program, which looks to the public – citizen inventors, academics, makers, artists, YOU – to help us advance technology and bring a different perspective to obstacles that gets us outside of our traditional solving community. See what else we’re working on here.

Make sure to follow us on Tumblr for your regular dose of space: http://nasa.tumblr.com