Washington DC IEEE-RAS 2008 Small Autonomous Robot Speedway Competitions

[vidam]

Dear Fellow Robotic Engineers,

I am very pleased to announce a series of Small Autonomous Robot Speedway Competitions sponsored by the Washington DC IEEE-RAS local chapter. This will take place this summer 2008. Please read further below for registration details.

Background
Robot competitions are entertaining and a means for evaluating the performance of mobile, autonomous robots for a given fitness evaluation. Ranging from mobile, autonomous wheeled vehicles such as the Grand DARPA Challenge to cooperative teams of humanoid robots that compete in soccer, such as RoboCup, the challenges are typically representative of the kind of engineering problems that extend the boundary in Artificial Intelligence, Computer Vision, and Automation. The fitness evaluation can be duration in time to complete a task, ability to complete a sequence of tasks to meet a goal, and ability of robots to handle uncertainty.

The Competition
The PIs have received a grant under the IEEE Robotics and Automation Society’s Local Chapter Grants for this purpose. It is our belief that competitions are an effective means of stimulating interest among participants by providing exciting technological problems to tackle. The Autonomous Small Robot Speedway competition is an outdoor, track-and-field race style competition. The track-and-field provides an excellent basis to measure the performance of small self-navigating robots in terms of speed, endurance, and sensor accuracy, and path planning. The rules and guidelines for the competition will be sent to interested participants.

Who Can Participate?
Under this effort, we are soliciting participants in the Greater Washington Area (Washington D.C., Northern Virginia and Baltimore). We would like to invite any IEEE junior members, University students, and the like to accept the challenge towards acquiring a deeper appreciation of the state-of-the art and the challenges that are currently the focus in robotics and automation.
The mobile robot challenge would begin this summer and there will be 3 tryouts before the final competition. The competition occurs once every month starting in June 2008. The robot size can range from 10x10x10 cm to 40x40x40 cm. The rules and guidelines for the competition will be provided to all the interested applicants by February 28th, 2008. First prize ~ $500.00 and vendor related prizes ~ second, third prize awards.

Interested?
Please contact us by April 30, 2008 with a succinct statement of how you expect to benefit from your participation and why you should be selected (emphasize skills and experience with respect to the requirements of the competition) to ieee.ras.dc@@@gmailDOTcom. If selected, you are expected to attend seminars, tryouts to be held at the University of Maryland over the course of the competition series.

Beyond the Competition …
It is our belief that this competition will serve as a model for establishing a university-community focused on a real-world practical problem. The proposed effort will be administered under the auspices of the IEEE Washington Section Robotics & Automation Society Chapter (Chair: Dr. Madhavan).

Rules:

http://ieee.ras.dc.googlepages.com/robotsspeedwayrules

Contact Details:
Melanie
ieee.ras.dc@@@gmailDOTcom

[vidam]

I just wanted to let everyone know the date to register has been extended till May 30th because rules have been updated due to a change in venue.

http://ieee.ras.dc.googlepages.com/robotsspeedwayrules

[Spoil9]

Sounds like fun. But I'll be in DC in May, not June.
Oh well, maybe next year.

[vidam]

Bonus Round:

Bonus round means it is optional. In the bonus round, your robot should recognize 1-2 typical
traffic signs and obey the traffic law. For example, one traffic sign
is a red stop sign like you see when you stop at a 4-way intersection.
Another traffic sign could be a speed limit that would be easily
attained by most robots. I'm thinking about how I would measure the
robot speed externally though.

I haven't figured out the prize for this round either. And I still
have a few logistics to workout such as where do I find the traffic
signs. I don't think it will be too hard though. Let me know if this
would be of interest to anyone.

The race is now Saturday September 6th and we have 4 teams participating.

-Melanie

[JesseWelling]

And I still have a few logistics to workout such as where do I find the traffic signs.

What do you mean? Traffic signs are everywhere... just borrow some 
But good work on the concept... I think it's a great idea for advancing robot theory.

[vidam]

Since my post, I was chatting briefly with a professor over at the University of Maryland about the bonus round idea. He has since then sketched an outdoor course along a sidewalk that includes traffic signs, a turn circle, and traffic cones. I will be posting an update to the website on the logistics for this part of the race if it can get approved with the administration staff.

Also, at the 2008 Intelligent Vehicle Transfer Technology at NIST a few months ago the goal was about getting automated robot vehicles into practical useful outdoor purposes in the future.
I think this bonus round competition is based alot on what I learned at the IGVT conference. If we can educate and advance robot theory towards getting automated vehicles into practical useful outdoor scenarios faster then this would fully relate to the goals of that conference as well as others.

-Melanie

[JesseWelling]

What is AutoTrac? Chopped Liver?

[vidam]

What is AutoTrac? Chopped Liver?

AutoTrac looks as if it can control a vehicle to go in straight lines in order to cut grass or plow a field. It is a bit more on the basic side than what I'm talking about here.  AutoTrac appears to use a programmed path.  It appears that it might not adapt to its situation. Without doing in-depth research of Auto Trac from first glance at the website ad, it looks as if it draws parallel lines and follows them.  What we ultimately want to do is make a robot recognize and respond to a changing situation, and this challenge is basically just a step in that direction.

[JesseWelling]

AutoTrac can do generally any preprogrammed path that is physically possible, and you are correct it doesn't respond to a changing situation other than keep the tractor accurately on that path. But it does have potential and I think it will really go somewhere someday. I'm just happy to work on something that has potential to become a full blow robotic project and was a little bit affronted when you stated your goal as "getting automated vehicles into practical useful outdoor scenarios faster" when that is exactly what John Deere is trying to do. Also see the R-Gator. I only wish I could get on that project.

[vidam]

All--

The details on the bonus round competition are online. Over the next couple of days we will post the GPS waypoints and images of key crossing points for this course.

Bonus Round: Cross-Campus Small Robot Autonomy Competition

Best,

Melanie

[vidam]

I've been using Google Earth to get way points of the campus. It is pretty easy to plot the map and get the way points yourself. Oh well, I guess I should at least post the *.kmz file for people who are too lazy to figure out the software.

[vidam]

I'm not sure if I should start a new post but I wanted to make sure this topic is associated with the original competition of this post.

The big news is that in the next few weeks, a simulation in Microsoft Robotics Studio of the IEEE-RAS-DC Robot Speedway competition will be publicly available. This is only for simulation and does not qualify as an official entry to the actual contest.
You must have a real robot at the competition to win and the deadline to register is passed. However, you could still play with the MSRS simulation code as it is fun and educational.

The current state of the image is what is shown in the image at the bottom. Right now the robot is a Pioneer ATV robot positioned in the center of the field. The robot is equipped with a camera that processes images orders the robot to move in a circle. The hope is that anyone interested in playing with the code could then write their own algorithms to process the images in OpenCV and move the robot inside the elliptical cones quickly and with the least penalties over a 1 km distance.

When the code is near release, I will put a link on this forum where you can download the code and try it out at home. There will also be some documentation and the code will be well-commented.



I would like to give proper credit to the programmers:
Dave VonHausen, Lalitha Duddey, and of course, me, Melanie.

[vidam]

We just launched a public website for the robotics competition. It is a work in progress. Feedback or suggestions on how the site could be improved would be nice! Thanks to Eric Martin of EM-Graphics for all his graphics work. Now he just has to finish his own site. 

[vidam]

Our competition is growing. We just picked up another sponsor from a regional Robotics company. YEAH!!!!! Jumping up and down for joy. I won't sleep tonight though.

[vidam]

Competition Event Guide Online: http://robotronics.org/images/guide.pdf
 

[vidam]

I just wanted to let everyone know that the competition is going to be rescheduled. There is a state of emergency in the State of Virginia because of this thunderstorm, Hanna.

The next date for the competition is October 4th.

[vidam]

We had the rain date on Oct. 4th. Weather was picture perfect and lots of fun. Three out of 7 teams competed. My robot wasn't anywhere near finished on time.    There is always next year. 

[youtube=425,350]dvgX6pmlulM[/youtube]

[youtube=425,350]UsB9j33obzM[/youtube]

[youtube=425,350]JNnyRjS_2WQ[/youtube]

[youtube=425,350]cH0DRISqvDw[/youtube]

[youtube=425,350]KLV5u7ikmHk[/youtube]

[youtube=425,350]GR_rDF6CDpk[/youtube]

The pictures are also online:

http://www.robotronics.org/pastcompetitions.html

The scores and team placements will be online soon.

-Melanie