Run Windows on Your Rocket!

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

OverlordvI

Well-Known Member
Joined
Mar 20, 2011
Messages
111
Reaction score
0
Has anyone played with the idea of running an Intel Atom processor on a COM express module to power the electronics bay of a rocket? There are some rather sophisticated USB-based altimeters, accelerometers, temperature sensors, video and camera devices that can readily be connected to such a platform.

The flexibility and ease of programming are also a must for someone who doesn't necessarily have a computer science / engineering background. Think about the things you could do -

With a "real" computer as the backbone, you can easily record high def video to an onboard solid state flash memory card, and stream it live via a Wi-Fi or Wi-Max connection to a receiving dish on the ground. The limit for such a transmission with the right (inexpensive) equipment can easily top 5 miles, and certain tests with clear LOS have found a distance over 50 miles.

Second, with high quality solid state sensors and components, plug and play into your payload bay is easy. Various modules can be integrated to control things like drogue and parachute deployment, electronic second stage ignition, A/V recording, telemetry, GPS, and more.

There are a ton of modules out there for communications as well. Setting up a radio beacon or a GPS locator becomes a matter of plugging in the right modules and making sure the software plays nice.

The super small computers are also very versatile in terms of environmental conditions, and since they are solid state, they can take very high Gs. With a properly supported payload bay, the somewhat high initial cost would quickly pay itself back over a couple dozen flights with high quality data.

The total cost easily scales with your budget. The smallest computers cost around $105, and the modules are all available separately. Their form factors are also very forgiving, with some of the smallest ones having a width of about 3" - and they can get smaller yet.

Power requirements are tiny since they use the Atom processor. Plus, you can turn off things that you don't need, like the VGA port.

The best part is that once the rocket lands, you can set up your data link wirelessly and download the hi-def recording of your launch via your laptop and wi-fi network.


At the least, I think it gives new meaning to the blue screen of death should Windows crash...

What do you folks think?
 
Lol, if you had Wi-max signal in your area plus wi-fi, you could turn your rocket into a Wi-Fi hotspot...
 
I have thought of using a Linux gum-stick size computer connected to a webcam to record video. Additional USB peripherals could be developed for transmitting data, GPS, altimeters, inertial measurement, and so on. The Linux gum-stick computers seem like a perfect match for rocketry.

Dave
 
Linux, of course, is the way to go. The problem is the programming issues - I can do HTML, but anything else is beyond my reach. A simple plug and play platform would be great.


BTW, using hte rocket for a Wi-Fi hotspot is a great idea :lol:
 
All the rockets could have a LAN party together as they're going up!:D:cool:
 
Didn't the Navy try Windows for it's shipboard computers back in the 80's?
If memory serves the pilot program ship's computers crashed and left the ship adrift.

The Aegis enabled ships systems are mostly Windows NT 4.0 based and yes, a BSOD would have rendered the ship defenseless and a bit blind (but never adrift). The last occurance of this BSOD was in 1998 on the Yorktown and there was hell to pay. Needless to say Microsoft was very motivated to fix the issue permanently. It's my understanding they've since migrated to a hardened Windows NT Embedded based platform. It's interesting to note that Aegis will pretty much run a ship by itself if necessary and even quite effectively defend itself against attacks without human intervention. Its rumored that it even has the ability to drive itself to a predetermined rendezvous point (defending itself along the way) if the crew were disabled somehow. I'd love to see the details current state of this system.
 
BK that's pretty neat. I didn't know that the AEGIS was so sophisticated.

In terms of real world BSOD, my experience has been that if you are working with quality components and experience no hardware blips or defects, you should not be having any BSODs with a clean and stripped version of Windows - the kind that would presumably power a rocket.

Of course, nothing about hardware is going to be blip-free when you put it inside a rocket.

The real question isn't necessarily about the software, as those bugs can be worked out, but whether you can reinforce the hardware enough so that the annoying little things like PCBs breaking in half from 10 Gs or a rough landing, or the SSD HD coming out because the rocket broke the sound barrier don't happen.


The potential for generating high quality data is very high with a good system. There are several very nice and portable high def 720p+ video cameras out there, starting for as low as $99 (Link 1, Link 2) that can interface with a tiny COM Express module.

The benefits are pretty obvious - first, you get solid state recording for basically an unlimited time (measured in hours, not seconds). You can hook up multiple cameras, so one is pointed down, another sideways, and a third looking up through the nose. You can record multiple sensory data, including temperature, air pressure, accelerometer information, GPS telemetry, and more. Setting up radio beacons and GPS transmitters becomes straightforward. You also get real time data acquisition.

The real benefit is that this high quality data is likely to be unique with the proper deployment. You could do some real research with information like this - for example, tie your computer in with real time data from a weather station, get permission to fly in a cloud during a rainstorm, and you can get very high quality data about temperature and pressure stratifications within a storm system. Combine that data with similar big clouds that don't produce rain and you can get a good leg up on figuring out what goes wrong in a drought even with big storm clouds passing through.

The hardware expense is variable, but a decent hardened system with high quality sensors, A/V equipment, and transmitters could be purchased for less than $1000, and more likely around $700. With the proper transmitters, you won't lose your rocket and your original investment.
 
Back
Top