Interactive Robotics Laboratory
- Learning by Doing

Copyright 2017 Yu Gu, WVU Interactive Robotics Laboratory. Contact Us


WVU  Statler College  MAE  LCSEE

NASA Sample Return Robot Centennial Challenge

The 2014 Story

We started very late in the game... SRR challenge has already been around for two years before we began to think about it. The good news was that we received great supports from the college, departments, and NASA WV Space Grant Consortium.
We were also able to build a team
with several outstanding students.

Everything had to move really fast.
It took us about six months from
conceptual design (December) to
finish the rover construction and
instrumentation (May). In the
meantime, we had to overcome
many different challenges, such as
fitting the robot through the door…

Things finally started to come
together in May. The electronics were
mounted; robot was moving,
computer vision algorithms started
seeing things. But we still had a long
way to go before getting every
individual part worked out, let alone
having them integrated. By May 8
a month before the challenge, we had to submit a video of a working robot. This is what we were able to come up with after some serious effort.


Sponsored by:

Benjamin M. Statler College of Engineering and Mineral Resources.

Grant Consortium:

MAE Dept.:

LCSEE Dept.:

We Gratefully Acknowledge the Generous Support Provided by:

The video deadline really exposed how much work we had left to do. Many team members started to spend 14-15 hours a day and 7 days a week working on the robot. Many problems were solved but more started to show up.

About two weeks before the challenge, things still looked really bad. The robot can not properly approach the sample, various devices were failing randomly, and computers were also kept crashing.  

To save the game, many simplifications were made. About a third of instruments and more than half of software steps were removed and every effort was made to make the system simpler and more reliable. The creativity of the entire team was stimulated and progresses were made daily. By June 7th, the day before us leaving for the Challenge, the robot was finally able to explore around, find the sample, pick it up, and bring it back to the starting platform. However, we didn't have enough time to do a full-scale test to see everything work together in a realistic environment. We had to leave…    

Everything in the lab was packed and transported to Worcester, MA with a box truck. At Worcester Polytechnic Institute (WPI), the host of SRR challenge, we were really happy to see a big enough lawn that for practice. Finally the robot got a chance to drive longer distance and explore multiple waypoints. However, the reliability issue was still hunting us and the computer crashed occasionally and speed controllers still failed once for a while.         

After successfully collected the
sample for 7 times, we decided
to call it done. Our robot was the
first to get inspected (to make
sure no illegal technology was
used) and first to be impounded.
Entering impound meaning nobody
from the team can touch the robot
(other than placing it on the starting
platform), and the autonomous robot will have to be on its own from this point on. With nothing else to do, we all left and patiently waited for the next day to come. Actually, we were the last one on the starting sequence, so our run was not until late afternoon of the following day.   

On Wednesday, all we had to do was to place the robot on the starting platform and walk to a viewing area. Everybody cheered when the robot started to move because there was a good chance that the computers might crash during startup... The robot did what it intended to do: drive off the platform; visit transition waypoints; visit search waypoints; scanned the camera and found the sample; and then collected sample. Everyone who watched this video knew what happened after that... So we walked away with lots of disappointments. But at least the robot picked up the sample, which earned us a $500 technological achievement award.

The good news was that the organizer gave everyone a second chance on Saturday. After more tests, we did some minor modifications (i.e. 2 numbers in the code); and it worked this time (see a video here!). You can see people's reactions after the robot brought the sample to the starting platform from the photo below. You can also see that some people reacted just a little bit slower than others :)

Everyone was happy. We were awarded $5,000 Level-1 prize money by NASA that night.  

The story has not ended yet. Please see our 2015 story here.

Robot: Our robot is named Cataglyphis, inspired by desert ant's extraordinary navigation capabilities.

Videos: A month before Challenge; First Attempt; Second Attempt.

Media: NASA, WVUToday, NASA 360, NASA Technology Innovation e-zine.