ICRA 2010 Conference in Alaska
The International Conference of Robotics and Automation is the biggest Robotics conference and it is held annually. This year, it was held at Anchorage, in Alaska. Why the organizers would choose a conference place that is far away from everybody is beyond me. However, it was an excellent chance to visit Anchorage, and a small bit of the state of Alaska.
I enjoy going to ICRA conferences mainly because of the large scale of topics that there are available to the attendees. I am interested in Surgical Robotics, and this year, there were two excellent workshops together with the normal tracks: A workshop on Meso-scale biomedical robotics, and a workshop on larger robotic systems (cyber-physical systems). I had a poster presentation on the first one, showing my results on "Ex-vivo experiments with Intraocular Microrobots", and attended the second one, where I was very lucky to see exciting presentations by Robert Howe, Greg Hager, Russel Taylor, Jacob Rossen, di Maio etc. Top that with a plenary talk by Blake Hannaford, and this year's ICRA had a lot to offer in my field of interest. I was unable to attend the Mobile Microrobotics workshop however, since it was overlapping with the Surgical Robotics one.
ICRA turned out pretty successful for our lab in general: Michael Kummer took the Best Manipulation paper award for his Octomag paper, Kathrin Peyer took the Best Student paper award for her work on characterizing the behavior of Artificial Bacterial Flagella, Brad Kratochvil had his video nominated for Best Video award, and lastly, we won 2 awards on the Microrobotic Challenge organized by NIST.
The Microrobotic Challenge was another particularly successful aspect of ICRA this year as well, in my opinion. The NIST guys (Craig McGray, and Jason Gorman) took the "nanogram" of the Robocup, and made it an event parallel to ICRA. It makes much more sense that way: more teams are able to participate, since they are anyway going to attend the ICRA conference, and the directly interested people find it easier to observe the advances in the research in that field, since they don't have to go to Robocup to see the new systems in action.
On the touristic side, I must say that it was pretty fun. It didn't reach the levels of the ICRA in Japan, but also in Japan we were only 4 people, and thus it was easier to organize ourselves. For one thing, Anchorage is a pretty nice city. It is relatively big, has the distinct US city style, where everything is spread out and you HAVE to have a car in order to even go and buy basic supplies, has some pretty nice bars, and cool night-life. The latter is to be expected, since a very big percentage of the population are actually young people doing their military service. Around Anchorage the nature is very nice, and there is coastal trail, the Tony Knowles coastal trail that takes you around the city. You can also see the occasional Moose around there :)
What is breathtaking in Alaska are the landscapes that you see when you venture outside of the cities. With Dominic Frutiger and Simone Schürle, we went to Seward, and Whittier at the south of Anchorage. Both are little villages that are surrounded by majestic mountains. It is super beautiful. We also took a cruise boat and took a trip of the Prince William Sound. We also saw a collapsing glacier!
What I missed to do was to spend a few days in the countryside, in a hut. Unfortunately I did not take any days off for this trip, so, as soon as the weekend after the conference was over, off I was for Zurich. That means that I need an excuse to go to Alaska again, and fill in the gaps :)

ICRA turned out pretty successful for our lab in general: Michael Kummer took the Best Manipulation paper award for his Octomag paper, Kathrin Peyer took the Best Student paper award for her work on characterizing the behavior of Artificial Bacterial Flagella, Brad Kratochvil had his video nominated for Best Video award, and lastly, we won 2 awards on the Microrobotic Challenge organized by NIST.
The Microrobotic Challenge was another particularly successful aspect of ICRA this year as well, in my opinion. The NIST guys (Craig McGray, and Jason Gorman) took the "nanogram" of the Robocup, and made it an event parallel to ICRA. It makes much more sense that way: more teams are able to participate, since they are anyway going to attend the ICRA conference, and the directly interested people find it easier to observe the advances in the research in that field, since they don't have to go to Robocup to see the new systems in action.



Compling Levmar
Levmar is a library that implements the Levenberg-Marquardt minimization. It has been developed by M. Lourakis at FORTH, Greece, it is optimized and widely used. I tend to run in trouble when I install it in linux everytime. In the last linux distros however, all the required libraries (i.e. libblas, liblapack) are easily accessible from the package manager. It took me a while to figure out that lf2c requires the package libf2c2, and libf2c2-dev as well. Moreover, liblevmar.a needs to be copied to /usr/local/lib, which should of course be in your LD_LIBRARY_PATH. Lastly, if it complains that it does not find MAIN__, just add
at the end of the demo file. This is some FORTRAN weirdness that comes into play I think.
void MAIN__() {}at the end of the demo file. This is some FORTRAN weirdness that comes into play I think.
cbrt in "math.h" (C++)
I am coding in C++ a pose estimator for devices that move in refractive optical systems... I need to solve 3rd order polynomial equations, which are straightforward to solve. However, in C++:
gives different results than:
which is actually a function in lmath that I didn't even know it existed.
And when I say different results, I mean that the 1st snippet gives NaN, whereas the 2nd the correct solution :)
UPDATE: Jesus guys, decide if you overload functions or not! According to that, pow is overloaded. What I see by compiling with g++ though, is that I need to use powf! I guess I screw up the includes. I read at other sites that I should include "cmath.h". Compiler grumbles that the file does not exist however. Doh...
pow(S, 1.0/3.0)gives different results than:
cbrt(S)which is actually a function in lmath that I didn't even know it existed.
And when I say different results, I mean that the 1st snippet gives NaN, whereas the 2nd the correct solution :)
UPDATE: Jesus guys, decide if you overload functions or not! According to that, pow is overloaded. What I see by compiling with g++ though, is that I need to use powf! I guess I screw up the includes. I read at other sites that I should include "cmath.h". Compiler grumbles that the file does not exist however. Doh...
When Grub can't find a device for /
Ok, so, here it goes:
I am using VirtualBox with Ubuntu 10.04 LTS, running on my host operating system Mac OS X. I use the guest system for development in C++, since all our robotics systems are running linux.
I usually put everything to sleep when I go home, including Ubuntu on the VirtualBox. All systems recover normally most of the times. Yesterday, however, I got this message while rebooting Ubuntu:
and I was into BusyBox for recovery instructions.
BusyBox does not allow you to do much, especially since /sbin is not mounted. Of course, you need your data back... I make sure that time machine backs up my virtual disks as well, but nonetheless, you want your data back...
The solution to the problem (after searching for a while in fora...) is to reinstall GRUB 2:
Now, at that point, number 6 did not work for me, but after restart, I got all my files back, and the system booted normally. Also, notice that after you mount your file system through the LiveCD, you can "scp" everything to wherever you want, thus being able to reinstall the system if the shit hits the fan.
I am using VirtualBox with Ubuntu 10.04 LTS, running on my host operating system Mac OS X. I use the guest system for development in C++, since all our robotics systems are running linux.
I usually put everything to sleep when I go home, including Ubuntu on the VirtualBox. All systems recover normally most of the times. Yesterday, however, I got this message while rebooting Ubuntu:
mount: mounting /sys on /root/sys failed: No such file or directory
mount: mounting /dev on /root/dev failed: No such file or directory
mount: mounting /sys on /root/sys failed: No such file or directory
mount: mounting /proc on /root/proc failed: No such file or directory
Target filesystem doesn't have /sbin/init.
No init found. Try passing init = bootarg.and I was into BusyBox for recovery instructions.
BusyBox does not allow you to do much, especially since /sbin is not mounted. Of course, you need your data back... I make sure that time machine backs up my virtual disks as well, but nonetheless, you want your data back...
The solution to the problem (after searching for a while in fora...) is to reinstall GRUB 2:
- Boot to the LiveCD Desktop
- Run "sudo fdisk -l"
- Run "sudo dlkid" to determine the particion that should be mounted (I assume there is no /boot particion). It should be something like sda1
- Run "sudo mount /dev/sda1 /mnt"
- Reinstall the GRUB 2 files by running "sudo grub-install --root-directory=/mtn/ /dev/sda" (notice the absence of the number there)
- Refresh the GRUB 2 menu with "sudo update-grub"
Now, at that point, number 6 did not work for me, but after restart, I got all my files back, and the system booted normally. Also, notice that after you mount your file system through the LiveCD, you can "scp" everything to wherever you want, thus being able to reinstall the system if the shit hits the fan.
Lago Di Como, Italy

The weekend of 27-28th of February, I visited the village of Como, at the Lake of Como, Italy. The drive there is absolutely fantastic! We passed through the village of San Bernardino, which is unfortunately not beautiful. Along the way though, it is hard not to be amazed by the beauty of the Alps.



All in all, it was a very sweet and relaxing short trip. I am not sure if I would want to go back to Como again, but visiting the lake is definitely something that I would do anytime.
IROS 2009 Conference in St. Louis

St. Louis is located at the border between the states of Misouri and Illinois, and is separated by the Mississippi river into Eastern St. Louis, and Western St. Louis. When one would cross the river, the Wild Wild West started. Since St. Louis is a city in the center of the US, it was a very famous station for people migrating from East to West. The city had a size of around a million in the older times, but has now diminished to around 300,000 inhabitants. St. Louis’ landmark is the “St. Louis Arch”, which is a 300 m arch that signifies the passage from East to West. Morever, St. Louis is famous for its jazz and ragtime music, as well as for the “St. Louis BBQ”.

