Thanks, the chassis was a rolling kit from Lynxmotion. My only complaint with it is the small size and I intend on building a larger custom chassis using only the tracks from Lynxmotion.
I thought as you did about the JavaScript scheme...ran a quick test this morning to give you a
general idea on the response time...I'm registering 3-4 key-press events per second. So, that's roughly
1/4 of a millisecond. You can tap keys and the tank jumps

I've yet to notice the firefox issue with the ram, maybe because I'm running 24Gb (I run multiple virtual machines simultaneously).

I dislike Java, Check out Scripting Layer for Android (SL4A); Python, Perl, JRuby, Lua, BeanShell, JavaScript, Tcl, and shell are currently supported. SLA4 exposes all the android API's and I found it a
joy to use. Maybe it will help.

I want to say I've seen a open source android control interface for the ArduCopter...
somewhere. Don't have the time to find it right now, gotta go make some money so I can
buy more equipment and feed my babies... Not prioritized in that order

Thanks for the info on the python package, I could have used a jump-drive and
extroot overlay but at the trade off of another usb port for the jump-drive wasn't worth it to me.

Keep me posted on your progress and let me know if you hit the proverbial "wall"
as problem solving is what I enjoy most!

Thanks again for your insight.
Funny thing is, I do have a GSM/GPRS Module in my parts bin
Security is one of my major concerns with the system I'm planning.
My plan is to have every light and all the electronic appliances controlled by my home server. I've started out with a pure multimedia solution, I have my TV hooked up to a small computer hidden inside an old HiFi stereo amp case. Through a webinterface on my server I can chose from a library of media files and DVB tuners that should be displayed on my TV. I can do that from all around the world -of course it doesn't make much sense to have something playing on your TV if you're not home... except maybe to fool potential thieves into thinking that somebody is home.

This basic Multimedia Automation System is going to be extended to a full home automation. Which means lots of microcontrollers that aren't powerfull enough to implementing WiFi like wireless security.
The Carambola on the other Hand would allow me to use WiFi security and send all the data through an asymmetrically encrypted TCP Stream (I've developed software for that purpose before). So even if someone hacks the wireless, he still won't be able to take over home automation.

But of course I'll still need small microcontrollers to control all the relays and lighting. So to not compromise security by using wireless transceivers I was thinking to use a simple form of the powerline modem - have the microcontrollers communicate with the Carambolas over the powerline -

Also, speaking about security, there need to be sensors connected to this network as well, so I can detect break in attempts.

Of course with a system this complex the ordinary light switch becomes obsolete, so they are going to be replaced by small touchscreen equipped embedded systems. At first I was thinking to build those myself with a small ARM Cortex-Mx microcontroller and those cheaply available touchscreens with 8/16 Bit interfaces. But then I stumbled across some 4.3" android powered mini tablet computers for just $45 a piece. Doing the math those aren't only cheaper but will offer more flexibility in programming. Since I own an Android phone and a normal sized Android tablet, Android will now be my primary platform for developing a remote control interface. (So your tip with the Android scripting engine is probably far more valuable than you previously realized )

Right now I'm developing one of the lighting solutions that will be integrated into the home automation network. It will contain a Fluorescent light and two RGB LED strips per unit. This is a low level unit, meaning it will contain an ordinary microcontroller. After it is built I can design a WiFi to Powerline node (Carambola based of course) and then start developing my home automation framework.

But before I start to integrate door openers, window blind controllers and other entry related components I want to make sure that the network I've built is as secure as possible. Just like building a Tank Drone before a flying drone, so you don't pay the highest price for a mistake.

Sorry for the delayed response, its been crazy on my end.
You've obviously done your homework and have a very solid plan.
I mentioned the security concern as so many ppl overlook it during the planning stages.

I like the powerline modem idea, as it allows the use of existing wire as a network infrastructure.
One exterior plug/outlet "could" allow physical access though, some thing to consider.
Schematics for a basic powerline modem are somewhere on my machine, but I've never got around
to ever building my own.

Your encryption scheme seems solid and well thought out.
I don't feel like you need any advice on network security...

PIR motion sensors indoors are a good choice if pets are not an issue.
They are cost effective, and cut down on false alarms that are normally
more prominent in video based motion detection.
However, I have found the "Motion" Linux utility to be among the best video based motion detection software,
if setup properly and running on a good machine.

Centralize as much as possible, don't forget about the breaker box!
You could strategically place a few micro-controllers here to handle most of
your switching needs. Adding in another box beside, above, or below would be
preferable to working in the existing enclosure.
Check local building/construction codes too, Where I'm at, we don't have any yet regarding residential automation.
A visit with the local fire marshal might be beneficial as well.

I've been running xbmc for my media-center a long, long time.
I'm curious now, what front end are you running?