It’s a structured-light 3D scanner developed by PrimeSense. While being used for measuring the three-dimensional shape of an object I think it should be possible to use it as a 3D vision system for indoor robots. So I would really like to see hackers taking this cheap device apart and re-purposing it for home-brew 3D vision. In their paper “A low cost 3D scanner based on structured light“, Rocchini et al. describe how to build one and this incredible instructables article explains how to build one also and even provides source code.

Of course there are other exiting technologies out there, a comprehensive list of 3D vision devices can be found at acroname.com (including prices). The devices range from expensive but effective and impressive Time-of-flight cameras (vendors: [1] [2] [3]) to LIDAR systems. The latter were very successfully used in DARPA’s Urban Challenge autonomous driving competition. A very interesting video that describes how LIDAR works in detail is available at blip.tv.

Of course there’s also the good old ultrasonic and infrared sensors that are cheap and have low resolution and are moderately reliable but they can certainly be very useful.

You can access the PDF from here and the source (OpenOffice Writer document) here. If you find it useful or would like an item added, feel free to leave a comment.

It’s a question, an error, a warning, an information and an idea icon. I currently use them in software documentation using the Shpinx Python Documentation Generator. It’s a very good tool to create beautiful documentation especially for software projects but I also use it to publish my game design document. Here’s an example of how the icons could be used:

I based the icons upon images from the Tango Icon Gallery. You may download an archive containing the source (Inkscape .svg) and PNGs of different size here. Use them as you wish, if you find them useful I’d be very happy if you left a comment here.

This was supposed to be a very literal copy of Master of Orion that you could play online with up to 15 other players on one map. Yet now I have something different in mind and I also have a different approach. It will only be a hobby project and I don’t plan to get immediate results, I want it to progress and evolve in my mind until I’m at a point where I can start writing code. Writing the code should be just handcraft at that point as I have most of the skills and knowledge to develop such a game.

What I’m doing now is writing a design document. The design will be highly influenced by the Foundation Series, one of the greatest science fiction series ever written. The first version of the design document can be found here. I still want the game to have a Master of Orion look and feel to it and it will certainly be a 4X type of game but I want different factions to be completely different (comparable to the different races in Starcraft) in the way they are played. I would like to see a “First Foundation” faction that relies on technology, economy and firepower and a “Second Foundation” faction that relies solely on the power of their minds, e.g. mind control. More about that can be found int the design document. A third faction is also possible but I have no concrete ideas about that yet. As the game design and the game itself progresses I will occasional write articles about it here on this blog.

Another caveat I found is also related to array indexes. When using a Matlab function that takes as parameters matrix indexes you might think of creating an mwArray of type mxUINT32_CLASS. Wrong. You have to pass indices – as strange as it sounds – as doubles so you should use mxDOUBLE_CLASS.

One custom Matlab function I wanted to use in C++ took as an argument a cell array. A cell array is an array containing other arrays. In Matlab you create a cell array like this:

a = magic(5);
b = ones(3);
z{1} = a;
z{2} = b;

If you look at z now you will get something like this:

z =
[5x5 double] [3x3 double]

An array of arrays. Easy. Well if you want to construct a cell array in C++ you have to do something like this:

mwArray a(5, 5, mxDOUBLE_CLASS);
mwArray b(3, 3, mxINT16_CLASS);
mwArray z(1, 2, mxCELL_CLASS);
z.Get(1,1).Set(a); // Sets z(1) = a
z.Get(1,2).Set(b); // Sets z(2) = b

You can find the description on this page.
If you get used to those little quirks however the Matlab Compiler is a very powerful tool to include your crazy Matlab scripts in a stand alone C++ application. One of the next things I would like to investigate are free (open source) Matlab Compiler alternatives such as the Octave C++ integration and compiler.

The two images show our drone dubbed B.I.R.D. (Basic intelligent research drone) in flight and the microcontrollers and sensors that control the drone. It is based on the the Mikrokopter project.

Another very interesting drone is the nano air vehicle by AeroVironment. It’s a flapping-wing vehicle for indoors use. Check out this amazing video to see it in action. I’m guessing that while the flapping-wing design is a lot more complicated than the brushless motors in our drone it is more energy efficient.

The third drone I would like to share was built by Armadillo Aerospace. They claim level 2 of the Northrop Grumman Lunar Lander Challenge with it. For more information read on here and be sure to check out this video. Notice how large this vehicle is (can be seen in the end of the video).

The fourth drone is really scary. And it’s not civilian at all. It’s called the Multiple Kill Vehicle and was supposed to shoot down intercontinental rockets in space. The project is apparently canceled but the video is impressive and scary nevertheless. Check it out here.

Update: Researchers at MIT’s Robust Robotics Group have developed a robotic helicopter capable of autonomously flying inside buildings or other GPS-denied environments. Here’s a very interesting video about their project. They achieved what we planned for our project – and I have to say it is very impressive and they did an excellent job.

I wanted to show multiple threads communicating and synchronizing with each other. I’ve read the argument that something like multithreading concerns only the implementation of a system and can thus be ignored when modeling. I agree with the argument however there will come a time in the development phase of a system when one needs to think about multithreading and one wants to sketch out how the system facilitates multiple threads to be efficient. And for that I have not found good tool within UML.

So I searched for examples of diagrams that visualize multithreading. The collage shows 4 different diagrams I found rather interesting. The first one is from this page about a traffic simulator written in Java (orignal image). It shows different threads and how they signal each other. It’s simple yet quite clear what is going an. The second image is taken from a CodeProject documentation about a genetic library (original image). It sort of uses UML symbols but it’s not clear what happens at the fork and join points. The way I figure it, those symbols inside the threads are supposed to be branches not forks. The third image is also from a CodeProject page – this time it’s a how-to about avoiding an exception when using multiple threads (original image). It reminds me of a sequence diagram. There are no blocking states in the diagram and I figure in this particular application the threads are non-blocking so diagram fits the application perfectly. The fourth diagram I liked the most. It’s from the Multithreaded Programming with ThreadMentor tutorial (original image) and has special symbols for sending signals and waiting for signals.

So I gave it a try myself and tried to model a simple controller and worker thread scenario based on the last of the example diagrams I found. The gray boxes and circle represent a signaling concept where one thread can send a signal and the other thread waits until it receives the signal. I also included a shared variable that handles the task shutdown. One thread can set the variable and the other branches depending on that variable. I am not happy yet with this particular part of the image. If anyone has a better idea or pointer to resources on that topic feel free to leave a comment.

We also found a quite professional solution in Funkamateur, edition 12/93, pages 794-796. The article is in German and can be found here. We found another very simple circuit diagram for an amplifier in Elektor, edition 7-8/2000 that we ended up building. Here’s the diagram:

The components cost about 10â?¬. We built in on two breadboards one for the amplifier itself and one for the power supply. With two 9V batteries connected it supplies -9V and +9V. Here’s a picture of the amplifier board:

Green red and black cables are connected to the power supply, the thick black cables are connected to the subject and at the bottom there are two pins to connect the oscilloscope probe or an AD-converter.
We used real ECG pads from the hospital to collect the signal from the subject. At first we used regular unshielded copper cables to connect the electrodes which resulted in a lot of noise. We then replaced the copper cables with shielded thin microphone cables and connected the shield to ground. This hugely reduced the signal noise. Here’s an image showing an EKG taken with our amplifier:

The electrodes are connected to right hand, left hand and right foot. As you can see it’s pretty much what you would expect. You can see the P-wave, the QRS-complex and the T-wave. We’re quite happy with the results.

If you try to build something like this yourself please be advised that you should be careful when connecting this amplifier to anything that carries high voltage (line voltage). The amplification circuit and measuring devices are not separated from each other. So you should only use battery powered devices like a USB oscilloscopes connected to a laptop running on battery. This is important as you consciously connect the subject to your circuit in a very well conducting manner (because that’s what you want) and at very dangerous points. So you should take care that no high voltages can flow between the electrodes at any time.

I didn’t have such a tool but it sounded like a fun challenge and I thought that such a tool might also help all the people who run into trouble when trying to create a spanning tree for their graphs. So I created walruscsv – a command line tool written in C++ that reads a CSV and outputs a LibSea graph file that can be visualized with Walrus. The CSV file hast to contain a link in each row and consist of 2 columns. The first column in each row represents the first node of the link and the second column the second node of the link. It automatically generates an arbitrary spanning tree.

A sample CSV file containing the links node1<->node2, node1<->node3, node3<->node4 could look like this:

node1,node2
node1,node3
node3,node4

Although Walrus is only capable of visualizing one network at a time walruscsv can handle multiple networks in one CSV file. It creates an arbitrary node and connects a node of each network to that node enabling Walrus to visualize the data. Currently the tool creates an arbitrary spanning tree so it’s not weighted and therefore not minimal.

Walruscsv reads the name of the CSV file to process as the first command line parameter passed to it. It is released under a non restrictive open source license so in addition to binaries the source code is provided. To compile it using the GNU C++ compiler simply type:

g++ wlink.cpp main.cpp wtree.cpp -o walrus.exe

You can use the program by typing

walrus.exe testfile1.csv

whereas testfile1.csv is the CSV file containing the graphs you want to visualize. Bloodshed Dev-C++ can also be used to compile the code and although I didn’t try it it should also work with C++ compilers from Microsoft (e.g. using Visual Studio). Here are the files:

Source code (the source code released to the public domain)
Test files (a collection of test CSV files)
Win32 binary
Debian x86 64bit binary

If you’re having trouble using or compiling the tool or if you have suggestions or comments of if you find a bug feel free to contact me or leave a comment.

About one and a half years ago I spent a considerable amount of time experimenting with social network visualization. My methods to gather and generate the data for the visualization tools I used were very limited.

However just after posting my first results I stumbled upon the Studivz Statistics page created by Hagen Fritsch who crawled the entire social network within only 4 hours in November 2006. Back then security measures were virtually non existent so it was an easy thing to do. Today the Studivz (which is a German Facebook clone) uses a Captcha to prevent such endeavors.

Hagen Fritsch was so kind as to make the data he crawled available to me. Thanks Hagen! Not all the data, just the relevant pages to determine the connections between users in particular the “Friends” pages. So now after 1 and a half years I found some time to look at the data, analyze it and feed it to Walrus. Some results:

The first graph consists of 17142 nodes and 22838 links whereas 17141 links are part of the spanning tree Walrus uses for rendering. The second graph consists of 25313 nodes and 37794 links. Four levels of the social network are displayed meaning the tree’s depth is 4.

This two images show one of the largest graphs I created: 413193 nodes and 1247702 links with a tree depth of four. The left images only shows the spanning tree and the image on the right shows the entire network. When working with such large networks I encountered a couple of problems.

I was using PHP to create the Walrus graph file. Because of the spanning tree this is no trivial task and it became very clear that PHP is the wrong tool for it. A PHP script ran for about 4 hours to create a 30000 node graph file. I wanted bigger trees so I ported the PHP script to C++ and was surprised to find that the same task took mere seconds. This is not due to inefficient coding but PHP definitely has problems when dealing with huge junks of data especially when they are stored in associative arrays. In C++ I used hash_maps which worked out perfectly. Here are two more images of the above network:

My Celeron 1.7GHz 512MB RAM workstation thought I was being ridiculous when I loaded the 60MB text file that contained the graph into Walrus. After some swapping and heavy calculation it worked reasonably well though.

If you feel like playing around with Walrus and the data here’s a zip file containing the 3 networks on display. You can get Walrus form here. Start Walrus with the following command:

java -Xmx512m -cp mp.jar;antlrall-mod.jar;libsea.jar;walrus.jar H3Main

I added the “-Xmx512m” part to allow java to allocate more memory. This is necessary for the very large graph file. Enjoy! I will try to add a video in the next couple of days. I was also able to extract the university each of the members of the social networks attends so there are a couple of possibilities to create more visualizations.