Embedded Linux Clustering (Part 1 - The Concept)

Linux clustering has been around for a number of years now. The
famous Beowolf cluster is the most well known example. Today there
are a number of technologies and projects that are developing Linux
clustering. With the use of LInux in more embedded devices, one of
the things that surprises me is that there is little to no research
being done on embedded Linux clusters. This article aims to address
this issue.

First off, I will start with a list of resources:
1. http://www.mosix.org/
MOSIX is a management system that allows a Linux cluster or an
organizational grid to perform like a single computer with multiple
processors.

2. http://openmosix.sourceforge.net/
openMosix is a Linux kernel extension for single-system image
clustering. This kernel extension turns a network of ordinary
computers into a supercomputer for Linux applications.

3. http://www.oreilly.com/catalog/highperlinuxc/index.html
This new guide covers everything you need to plan, build, and deploy a
high-performance Linux cluster. You'll learn about planning, hardware
choices, bulk installation of Linux on multiple systems, and other
basic considerations. Learn about the major free software projects and
how to choose those that are most helpful to new cluster
administrators and programmers. Guidelines for debugging, profiling,
performance tuning, and managing jobs from multiple users round out
this immensely useful book.

4. http://lcic.org/
This page is intended as a central repository of links and information
regarding Linux clustering, in all its forms.

5. http://oscar.openclustergroup.org/
OSCAR (Open Source Cluster Application Resources) is a snapshot of the
best known methods for building, programming, and using HPC clusters.
It consists of a fully integrated and easy to install software bundle
designed for high performance cluster computing. Everything needed to
install, build, maintain, and use a Linux cluster is included in the
suite, making it unnecessary to download or even install any
individual software packages on your cluster.

The Brains of the Robot - Embedded Linux

Through some research and some stumbling, I have come across are really nice micro computer capable of running embedded Linux that would make a great platform for the "brains" of this robot project. Like all new computer projects, this one will run embedded Linux. The gumstix has a very small, very capable platform from which to start.

http://www.gumstix.com: This is the commercial site from which to purchase the base gumstix platform.

http://gumstix.com/oscommerce-2.2ms2/catalog/product_info.php?cPath=26&products_id=81&osCsid=fd3fef04fb85223c295c0810200e235e: This is the most desirable platform.

Adding UWB MIR Radar System to my Robot

Ultra Wide Band (UWB) and Micropower Impulse Radar (MIR). These are the items I have been researching for the past few days. I have learned a number of things about MIR, and have jumped to alot of conclusions about what I could use this technology for on my robot. This article will atempt to contain my thoughts on the subject (for now) as well as illustrate some interesting (and possibly unrealizable with current technologies) ideas for how this technology could be used with a personal robot assistant (PRA).

Micropower Impulse Radar (MIR) was invented and developed by Thomas McEwan while he was at Lawrence Livermore National Laboratories (LLNL). Thomas McEwan has sinced started a company with this technology and the following URL will have additional information: http://www.mcewantechnologies.com. Following this link will show that the technology is available for lincensing.

Another interesting site for an example of how this UWB radar is used is at the following URL: http://www.timedomain.com/. Here the technology is used to see through walls to provide a tactical advantage. The various videos show police officers and/or soldiers demonstrating the product. Quite impressive.

The orignal LLNL site (http://www.llnl.gov/IPandC/technology/profile/sensor/MicropowerImpulseRadar/) gives a good overview of the technology. This URL fro mthe department of energy gives more ideas for how the technology can be used: http://www.eurekalert.org/features/doe/2004-09/dlnl-etu091604.php.

While the idea of having a visioning system on my home made robot that could peer through wals seems like a good idea, I am thinking of a different use for the technology. This technology can be used to see through skin and bone to get an image of the heart, lungs and digestive system. If such a tool was available in a hand held unit, doctors at the emergency room could have an extremely valuable tool for triaging a patent. Adding such a technology to a personal robot assistent (PRA) along with some rudimentary analysis of the resulting data could provide a rapid home diagnosis and alert system for an individual with a chronic alment. The data could be streamed directly to a personal doctor for analysis if the data failed to follow a normal pattern. The robot could also be programmed to search the premises for a particular individual in order to perform a scan. Such a robotic system could be used in hospitals to check on patients during the middle of the night or when the nursing staff is too busy.

This medial use of UWB MIR requires alot more research on my part, so here are a few more bookmarked websites:
http://www.econ-pol.unisi.it/high_frequency/08-staderini_abst.pdf
A paper on using the technology to remotely monitor heart rate (no electrodes needed).

More research to follow...

Developing my Own Robot - Part 1 The Concept

Ever since I was a kid, I have always wanted my own robot. Something similar to the Lost in Space robot that could do anything. Now that I am grown up and have children of my own, I realize that building a robot means many things to many people. In this article, I want to specify the desired features and functions of my robot.

The first feature of the robot is locomotion. I want this robot to be able to move around on a hard wood floor. This movement will on top of two roller skate wheels, with a trailing wheel for stability. The reason I specify this exact arrangement is that I have already purchased a round aluminum platform for the robot complete with roller skate wheels and servo motors. This is the platform that I am starting from: http://mivasecure.abac.com/bpatton/merchant.mvc?Screen=PROD&Store_Code=RS&Product_Code=MBB&Category_Code=MFSZ. Here is the main website for neat robot parts: http://www.robodyssey.com/index.htm.

The next feature that I am looking for in this robot is vision. Here I am undecided on the nature of the vision or on the technology. The simplest form of vision would be a computer camera. However, seeing in the dark using infrared lights would be a nice feature. An incredibly cool feture would be a micro radar, but it does not appear that this technology is ready for the hobbiest...yet.

The third most important feature is web based control. I want to be able to go to my home's website and control this robot as well as see what the robot is seeing. This third feature brings up the need for the next feature.

Wireless control is the fourth feature. Since I want to control this robot (which will be downstairs) from my web server (which is currently located upstairs) I will need wireless control. This feature will come from the numerous RC cars, trucks and tanks that me and my kids play with. I already have in mind one of the cars that I can part out for this feature. However, it would be much more desireable to simply add a wireless network card. I do happen to have an 802.11g pcmcia card sitting unused fro mmy old laptop. This card could provide the wireless capability as well as the live streaming video feed. Yes it is decided, I will use 802.11g for wireless connectivity.

The fifth feature is a processor with the capability to run embedded Linux. I want to have Linux running on the robot so that I can take advantage of all of the really cool software packages being developed for robots running Linux. Here is a website where I found a few of these packages: http://www.aaai.org/AITopics/html/soft.html.

The sixth feature that I want from this robot is audio. The robot needs to send and receive audio. I envision this being used to communicate with someone in the same room as the robot from a person connected over the web interface. A great example of this would be the grandparents wanting to talk to us or the children. They get on the web interface, log in, then move the robot around until they see one of us. They can then start talking. We won't see them, but we will see the robot. My youngest child (currently unborn and due 2/16/2005) would probably thing grandma and grandpa ARE the robot.

The final (until I think of more) feature is the ability for the robot to detect when the batteries needs charging and to automatically find its recharging station and connect. This is a feature that I thought of when I was looking at the first versions of the robomower (http://www.friendlyrobotics.com/robomo.htm) and the roomba (http://www.irobot.com/consumer/). These now both have this feature, and I agree that it is something required. Who wants to have to worry about a robot being charged up to play with it - it should always just work. I have this same complaint about all of the remote control cars, they have to be charged.

That completes my wish list of features. I will continue this article once I have purchased more parts of decided upon the next step of the project.

Map Websites

I found some very incredible web sites that allow you to build your own map and have it printed out and mailed to you. Since I recently bought a canoe (and picked it up tonight) I have been looking for a good map of the local area lake (Jordan Lake). The following is a list of websites and what I thought of their maps for sale.

http://www.mytopo.com/
This website allows you to zoom in and out, pick any place in the US and purchase several size maps. These maps can be water proof or glossy and you can order the following sizes: 36"x44", 24"x36", 18"x24".

http://www.fishinggpsmaps.com/
I found some cool maps here. These maps are especially cool if you have a GPS.

http://www.omnimap.com/catalog/cats/fish/fish-nc.htm
Talk about cool! These maps show fishing hot spots.

Building my own external USB DVD-RW Burner

I recently built my own external DVD-RW burner from spare parts that I had around. The first and most important part that I had around was - suprise - an EIDE DVD burner from Sony (the 500u internal). This burner was sitting in a computer that died several months ago and which I had not found the time to rebuild. Most of my computer time is now spent on my fairly new IBM Thinkpad T41p - which is a really nice laptop. Since I still wanted to burn DVDs (for personal and work), I needed to get an external DVD burner that connected to my laptop. Since I am cheap, I didn't want to spend alot of money. Shopping around at BestBuy, I could not find any USB 2.0 to EIDE cases. Having a $50 gift card for BestBuy, the next best thing was a USB 2.0 CDRW burner which could burn at 52x. I bought this CDRW burner (which had a very attractive black case).

The first thing I did was to connect the USB burner to my USB trace tool (a CATC Advisory capable of capturing USB 2.0 traffic). I was excited to learn that the protocol which the USB burner uses is the SCSI class and bulk-only subclass (I develop a remote drive solution for IBM and chose this very same protocol for the USB CD drives that I emulate over a network). This new toy would provide a very comparison tool to the stuff I develop.

Well, I opened up the mechanical package of the CDRW burner, managed to pull out the CDRW drive (which was a standard EIDE drive like I suspected) and cram my old DVDRW burner into the case. Using a little electrical tape (since the screw holes didn't match up exactly) I put the drive case back together with the EIDE DVD burner in place of the CD burner. The first burn I tried was at 4x which failed with a buffer under run. A minor setup. The 1x burn worked perfectly. I plan on a 2.4x burn soon. Having a 1x external DVD burner was definately more usefull to me then a 4x DVD burner sitting in a dead computer. Overall, a job well done.