Note from the team lead…

The Georgia Tech-SAIC Sting Racing Team has been informed that we were among a set of participants that did not advance to the Urban Challenge finals on November 3rd. As a first-time entrant, the team has done an outstanding job making it to the semi-final round of the world’s most challenging robotics competition. We wish to express our gratitude for all the support we have received from Georgia Tech and the Atlanta community, as well as the expertise and technical support from our industry teammate, SAIC. We also give a hats-off to DARPA for administering a fair and flexible race.

Looking forward, we are excited to continue our part in the evolution of autonomous robotics research and the development of intelligent, safe vehicles. Our efforts in the DARPA Urban Challenge – particularly our autonomous vehicle, Sting 1 – will play a key role in our future research agenda, particularly as we study the safety and viability of implementing robotic elements on the road and in the home. We will remain here in California to root for the remaining competitors and learn all we can for next time.

Tucker Balch, Team Lead
Georgia Tech-SAIC Sting Racing Team

Sting Racing is no longer in the race

Last night we were informed by DARPA that Sting 1 will not be able to qualify for the finals. They requested that our race slots be used for teams that might make the race. Obviously we agreed on the schedule changes to accommodate teams that might make the finals.

It has been a race with some serious challenges after the crash last Saturday. We have worked hard to rectify the problems and to address the unforeseen problems that we have encountered. Unfortunately it was not enough.

The team is obviously very disappointed. We have worked hard to get to the race and have a truly amazing platform. We are now going to watch the rest of the qualifiers and the finals to learn as much as possible. In addition the team will take it easy a few days to recover from weeks of long hours.

Our run in area C

youtube.com video

Monday PM: Sting Racing bounces back!

We’re very happy with our run on Course C today. In this course the vehicle runs loop after loop in a housing area, and other cars (driven by stunt drivers) interact with it at four-way stops. The robot got stuck once, requiring an intervention — our lead programmer had to restart one of our processes. We’re still assessing the cause of the problem. But overall the vehicle behaved very nicely.

Tomorrow morning we’re back on in Course A. This is the one that is grinding up lots of cars — besides our accident, there have been several others there. We’ve been investing a lot of time tuning our code for that test.

Some movies from the Urban Challenge

Our E-stop test and first autonomous run.

Visit the field station.

Sunday: Competing again

We were able to completely rebuild our sensor package, calibrate the sensors, and test the system Sunday morning.

We ran Test B in the afternoon.

Saturday PM: Sting Suffers Setback

In our Saturday afternoon test in Area A Sting 1 crashed head-on into a concrete barrier. The front sensor mount was bent severely and pushed into the front of the vehicle. Fortunately the protection offered by the design and strength of the sensor mount saved the sensors.

After assessing the log data we determined the cause of the accident was a failure in the communication link between our GPS/IMU and the main control computer. Without “pose” information the robot could not know that it was moving.

Recovery: Tartan Racing helps out

Our first order of business after the accident was to assess the damage and develop a plan, and we needed to fix the problem that caused the accident in the first place. The GM mechanics with Tartan Racing (CMU & GM) came well equipped to the site, and were eager to help us out. With their help we cut the main sensor mounting plate out, bent the metal back into alignment and welded everything back in place.

We give a big THANK YOU to GM and Tartan Racing!

The loss of comms between the GPS and the control computer could have occurred in several places, including the computer’s hardware, the linux serial drivers, or our code. We have circumvented the first two possibilities by installing a serial to USB converter and we access the GPS via USB. On the software side we added checks to make sure that pose information is streaming in, otherwise the robot brakes and reboots the computer.

Back in the game

After yesterday’s unfortunate afternoon collision, we’re back in the NQE. Around 9pm last night we were able to get things straightened out and recalibrate in time for this afternoon’s next test. DARPA says we can re-do yesterday’s tests as well, which is great. Many thanks to our friends at CMU and Stanford for being so supportive and helping us repair the vehicle. More to come…

The NQE has started

It is official. The National Qualifying Event for the DARPA Urban Grand Challenge is underway. We are back from the opening ceremony (had to be here 6:30 AM, left the hotel at 5 AM:).

Today the team has to perform safety and e-stop tests, and with time over in our timeslot (30 minutes) we will perform a second practice test in area 1. In addition some final software checks will be performed. Tomorrow the team will perform live tests in area C (morning) and A (afternoon).

Practice testing for NQE

Today STING had an opportunity to practice in order to evaluate the system in autonomous mode after the shipping to Victorville, CA. The allocated practice period was only 30 minutes. The areas for practice are two dirt areas, so things like lane tracking and road detection are not that easy to test However, the areas are great for assessment of the overall sanity of the system.

The practice period included setup of a new Road Network Definition File (RNDF), evaluation that all sensors are delivering expected information and a test of the newly generated RNDF. After this the vehicle was requested to complete the mini course in autonomous mode with no one in the car. As expected the vehicle did well and executed without problems the mission (1.5 laps on the oval shaped course).

Finally we also tested that the vehicle can be paused and resumed from the government issued e-stop console. Again the vehicle performed as expected.

There was an offocial welcome meeting with DARPA. The schedule has shifted a bit and the idea of two separate phases at the NQE has changed. Now navigation and traffic handling has been combined into joint tests. The actual testing has bene postponed by half a day to start on Saturday morning. Friday is reserved for practice testing, safety inspection and e-stop verificiation. The race is still 3rd November.

Yesterday (Wednesday) the skyes in the Mojave desert were blue and pleasant. When we drove home the skies towards LA were also clear. Today the winds had shifted back, but were not as strong and the smoke from the forest fires were back. There is a constant sense of smoke but at least we have not seen mcuh ashes the last few couple of days. It is a bit chilly in the morning (64F) but during the day the temperature increases to a comfortable 80F. It is however very dry in the desert.