Ingo Lütkebohle's blog

One of least known, but most important, aspects of the ActiveMemory is its support for indexes. The first, but actually less important, reason to use indexes is performance: As in conventional databases, indexes are essential for good performance. Due to the simple nature of the XPath expressions used in AM queries, performance is still OK without them, particularly when the Memory has only a few entries -- as is usually the case when people test their components, so this often goes unnoticed.

However, the more important reason to use indexes is that, without indexes, deadlock potential is much higher.

For a couple of years now, general-purpose computation on the gpu (GPGPU) has been all the rage to get the last bit of performance for many data-intensive algorithms, including computer vision on the gpu. Claims of 100-fold speed-ups are not unheard of and 10-fold is pretty common. Now, a couple of researchers from Intel have published a

Most current mobile robots use a laser scanner for building maps and these things cost serious money. Therefore, when Neato built a laser scanner into a vacuum for <400$ and claimed the scanner cost less than <30$ many people became interested very fast.

Yesterday, Willow Garage officially celebrated the start of their PR2 beta program, where they lend PR2 robots to eleven (11) institutions, ten of which are universities, with the eleventh going to Bosch. After the usual speechs, the PR2s made their entrance, where they were being led by the hand (i.e. physical HRI) by the Willow Garage team.

I'm currently visiting Willow Garage and a big reason for that is discussion about the design ideas behind ROS, the Robot Operating System, which has been initiated here. ROS has two parts: A core, which provides naming and messaging, and a large and growing number of extension modules for various typical functions such as navigation, control, etc.

Christian just pointed out to me that PtGrey's most recent newsletter has a nice intro to the upcoming USB 3.0 interface standard for cameras. Towards the end there is also a nice comparison to the various other interfaces with a good view of pros and cons.

OpenCV, the open-source computer vision library originally sponsored by Intel and now by Willow Garage, has been adding functionality at an amazing pace for the past 18 months or so. Much of the new functionality plugs holes that are particularly interesting for robotics, e.g. the new stereo pipeline, or adressing developer usabillity (e.g. the C++ and Python APIs).

While doing some diagrams, and checking out tools to do so, I am reminded of a saying attributed to Andrew S. Tanenbaum "The nice thing about standards is that there are so many to choose from".

Really, there are some diagramming standards I now encounter that I have never even remotely heard of before, and probably rightly so. A small selection, to date:

On Tuesday I finally put up the talk slides from the EBS-RO09 workshop. This has taken a while, as upon returning from the IROS, I immediately buried myself in the preparations for our final project demo in early November. And now, of course, its writing time ;-)

These guys obviously had fun doing their paper. I am usually sceptical about the real-world applicability of computer vision algorithms, but I can immediately see the value their algorithm has for geeks who regularly tele-conference or work from home. Check it out and you'll see what I mean ;-)