Carpe Noctem Robotic Soccer

We had our first game yesterday, CoPS vs. Carpe Noctem. It started late because of the usual delays at 19:15. The first half was not so bad even though we noticed very strange behavior of our robots. The first half ended 0:2 for us.

We tried to find the problem why our robots acted strange and even stopped randomly. It could not be fixed in the half-time, so we had to send them in the second half without much improvement. Finally the game ended 4:2 for the CoPS because at some point in the second half every robot of our team stood still.

After a sleepless night for many of our team we finally know that the problem lies most likely in a too expensive communication and old bugs that should not be there. We think that our SVN must be faulty because we found old bugs in different modules that were fixed and tested on at least two different machines where one checked out the fix from our repository.

Our team worked hard that night so the next game will look better for us.

The next game is scheduled for 10:00 today, but expect delays :-)
You can watch the field from our webcam.

Our webcam of the MSL-Field is online now and can be watched here.

Edit:
The Hannover-Messe router of our hall has severe load issues. If the image is not updated the router is most likely down again.

Long time has gone since the last post and this is the first post from the German Open Open 2008 in Hannover. The last two days were quite stressful and an emotional roller coaster. Nothing was really finished or working, a camera stopped working, and the motion of one of our robots stroke a light.

Nevertheless, more or less all robots are now up and running. Today there will be scheduled training games from 1pm to 6pm. In the morning teams may set up their equipment. Ten teams were registered by only eight attended the competition. We thus will have a quite relaxed schedule with no more than six matches a day. For Tuesday and Wednesday the following schedule is applied:

Edit: New schedule! (see also http://robocup.das-lab.net/)

Tuesday (updated 2008/04/21, 23:19)

09:00 Mostly Harmles — Ulm Sparrows
10:00 CoPS Stuttgart — Paderkicker
11:00 Brainstormers Tribots — Tech United
12:00 Lunch break
14:00 Paderkicker — ISePorto
15:00 Ulm Sparrows — Brainstormers Tribots
16:00 CoPS Stuttgart — Carpe Noctem

Wednesday (updated 2008/04/21, 23:19)

09:00 Brainstormers Tribots — Mostly Harmless
10:00 Carpe Noctem — Paderkicker
11:00 Ulm Sparrows — Tech United
12:00 Lunch break
14:00 ISePorto — CoPS Stuttgart
15:00 Tech United — Mostly Harmless
16:00 ISePorto — Carpe Noctem

We’re currently preparing for the RoboCup German Open. Till was asked to write a kicking behaviour to make the most of our new kicking device. The following happened this afternoon when testing and calibrating the kicker:

Further video formats:

• MP4: TellPAL02.mp4
• Flash: TellPAL02.flv

We’ve prepared a webpage containing organisational information for the MSL league:

http://robocup.das-lab.net/

Hendrik (he does not have a nick yet) is responsible for our new motion with an integrated stabilization and antilock break system. This performance was captured by Elm just about an hour ago:

Further video formats:

• MJPEG: cn2008-stunt.avi
• MP4: cn2008-stunt-mp4.avi
• Flash: cn2008-stunt.flv

The LabCam is up and running for this team weekend again:

http://www.vs.uni-kassel.de/~phbaer/cam/

Update:

There’s yet another LiveCam fresh from the playground:

http://www.uni-kassel.de/~flseute/livecam.jpg

Once again we’ve pitched our tents in the canteen for a weekend. There some new features to test and also new hardware.

The LabCam is up and running, providing you with the hottest … well, yes. Here we go: http://www.vs.uni-kassel.de/~phbaer/cam/.

Christmas staff party

We’re organizing a Carpe Noctem Christmas party next week. All members, fans, and supporters are welcome:

Carpe Noctem Robotic Soccer — DAS-Lab (Room 1405)
Distributed Systems Group
Kassel University
Wilhelmshöher Allee 73
34121 Kassel

We used the last weekend primarily for testing our learning framework (developed by puto). As a side-effect, we had also time to test the new achievements in out vision system (on Monday morning, 4-6am). One very impressing feature is the new segmentation filter which makes calibration almost dispensable. We were thus able to play with three different balls, one of them begin a standard black-white human soccer ball. Without talking too much, just have a look yourself:

Further video formats:

• MP4: cn2008-cibt.avi
• Flash: cn2008-cibt.flv

We’ve moved our field to the canteen this weekend to test some leaning algorithms. Please find our temporary labcam here: http://www.vs.uni-kassel.de/~phbaer/cam/.