[robocup-nao] Dissertation: Stochastic filtering on mobile devices in complex dynamic environments

[Stefan Tasse]

Hi everybody,

a friend reminded me that I forgot to send this mail so far, so better 
late than never:

I recently finished my PhD thesis and I would like to share it. I know 
most of you might stop reading if I copy&paste my abstract, so I will 
keep it even shorter:   ;-)
The dissertation describes some cool ways to efficiently handle sparse 
and noisy sensor data for reliable localization of robots and tracking 
of dynamic elements in their environment. It involves RoboCup as well as 
the application of the same concepts to other scenarios like 
intra-logistics and indoor smart phone localization. Building a 
localization system is easy, but doing it the right way will improve a 
lot in terms of accuracy and robustness. I highlight some of the things 
which are really worth thinking about when doing this, and these are not 
necessarily the ones you find the most publications about when googling 
"Kalmen filter".

The thesis can be downloaded for free at the TU Dortmund website:
http://hdl.handle.net/2003/33007

For those who don't know me:
I have been involved in RoboCup since 2005, in the 4LL, HL und mostly 
the SPL. I naturally did some research in different robotics areas 
(biped walking, machine learning, etc), but my focus is on sensor 
processing and stochastic estimation.

Best regards,
Stefan Tasse