Monday, January 25, 2010

Near-Space Photography Attempts 1 and 2

Yes, you read that right. I'm experimenting with high-altitude amateur photography. I would have posted about it before, but I wasn't sure this was actually going to happen until last week. Here's the story:

Around a year ago, I read an article on the BBC News website (which I highly recommend) that described a way to make a spacecraft with close to the same capabilities of Sputnik using household items. I was fascinated by the concept of building an actual spacecraft by digging up junk around the house, and I briefly considered doing it. However, I'm fairly results-oriented, and because I didn't have a way of launching the resulting spacecraft, I wasn't really interested in putting the work into building it. Besides, I was arms-deep at the time building my second eight-foot light bucket, which I'll talk about in another post.

Then four months ago, I was contacted by a high-schooler named Megan who had found my name on an astronomical website. She wanted help building a simulated satellite as a science project. This satellite would take pictures and transmit them to the ground during flight. I was initially skeptical due to the fact that launching even a small device would take far more rocketry than I (and indeed most people) had the expertise or budget to handle. When she told me she wanted to use a weather balloon to lift it to the upper atmosphere, I thought that this might just be possible.

Around the same time, I read about a group of MIT students called 1337Arts who had successfully done something similar; they lifted a camera attached to a weather balloon 18 miles high, took pictures, then retrieved them when the craft landed. They did it on a budget of $150 and appropriately called it Project Icarus. That's when I knew this would be possible. And as a bonus, they mentioned on their website that FAA regulations state that you can launch a free-floating balloon without any kind of clearance so long as it is under four pounds (along with some other minor restrictions).

However, our task was a little more difficult than Project Icarus. We wanted not only to take pictures, but to transmit them to the ground during the flight. Project Icarus transmitted GPS data using a cell phone to track their craft's location, but I couldn't find a way to get a cell phone to transmit pictures along with the GPS data with limited bandwidth and the kind of control we needed. The solution we came up with was to get a lightweight laptop computer, attach a webcam, a GPS receiver, and a wireless internet adapter to it, and launch the whole works into the atmosphere. As a result, we may be the first amateurs to attempt to launch a laptop computer into the upper atmosphere. I haven't been able to find another example, but if you know of one, please post it in the comments.

The laptop chosen was an Acer netbook with Windows Vista. Using Vista wasn't an ideal setup, but the laptop came with it and Megan was using it for schoolwork. I didn't want to cause her problems by installing XP or Linux, so I decided just to work with what was already there. It weighed less than two pounds and had enough battery life and hard drive space to last the entire flight. We used a weather balloon and parachute purchased from a scientific supply website, and we took 1337Arts' idea of using a styrofoam cooler and hand-warmer packs to keep the components warm during the flight. We used a Sprint Wireless Broadband adapter so that we could transmit pictures and GPS data to cell towers, and we used basic sub-$30 USB GPS receiver and webcam models. We also used a modified BeepX rocket retrieval beeper to make retrieval easier once it landed.

The report is due Wednesday, and we just started setting it all up last week, so it was sort of a frantic weekend. I have a Xubuntu server at home running my security cameras, and I had set it up so that I could access the camera images from my phone. We decided that it would be an ideal way to store the laptop's images and GPS data as well, since it would allow us to access everything from my phone as we were chasing the balloon without dealing with the cost and hassles of a third-party service. So I set up my server to handle the task on Friday, then on Saturday, Megan and her mother brought the laptop to my house so that I could set up its software.

I used completely free software for the project, making this an inexpensive way to perform the experiment. I installed GPS TrackMaker to record the GPS data from the USB device to a file that was updated every 30 seconds, and I set up Yawcam (which I think Megan had installed) to save images every ten seconds for later transmission to the ground. I decided to go with cwRsync to load the data onto the server using an SSH connection. cwRsync is a piece of software that synchronizes a local folder with one on a remote server. I then set up a Windows scheduled task to run cwRsync every five minutes. The advantage of using this setup for data transmission was that if the balloon left Sprint's coverage area, the computer would start from scratch every five minutes to attempt to send new data to the ground, so there was much less chance of a permanent failure due to a temporarily lost signal. Also, cwRsync allowed the computer to send only new files every five minutes instead of sending all of them every time, so it cut down significantly on bandwidth requirements.

The software took about four hours to setup properly, though with the information I now have (most of which I presented above), it would only take me an hour to set it up if and when I attempt it again. When the software setup was complete, the three of us drove out to the countryside to test poor and lost signal conditions. There were a couple of minor glitches, but for the most part, it worked beautifully. We were almost ready for the real thing.

That evening, Megan and her mother assembled the craft, and then they picked me up at 6:00 the next morning. We checked the University of Wyoming's balloon trajectory forecast site, and decided to launch from a coastal town about two hours away from us, which the site predicted would cause the balloon to land within 30 miles of our homes. When we got there, it took a couple of hours for final assembly of the craft and to work out the last few bugs in the software, but we were finally ready to go. Here is what happened:



That's right, complete failure. We actually ran out of helium when filling the balloon, even though we were told that we would only need about a quarter of what was in the tank. We discovered later on that this was because the people who had helped us calculate how much helium we would need had misplaced the decimal point. We were wold we needed to use about 250 pounds of tanked pressure when we actually needed closer to 2500. There was only 1000 pounds in the tank. To make it worse, the town we were in didn't appear to have any helium suppliers. So we had to deflate the balloon and go home with the empty tank. Then an accident occurred on the 2-lane highway between towns only 20 minutes from home, and it completely blocked all traffic, so we had to drive most of the way back to the coast to take the other highway. That added over an hour to our trip. And on top it all off, it was pouring rain the entire time. Still, our spirits were surprisingly high when we got back and we decided to try it again.

It was 1 PM by the time we were ready to give it a second shot. We didn't have enough daylight left to go back to the coast and still have a chance of retrieval before nightfall, so we decided to throw caution to the wind and launch right from the parking lot of the party supply place where we were getting our helium. We calculated that this would put the balloon in the mountains in very rough terrain, but there wasn't anything in the craft that we weren't willing to lose, and we wouldn't have time to try again another day, so that really was our best option. Here's what happened this time:



Forgive me for getting a little giddy in that video. See how much bigger the balloon is this time? Yeah, that helped. The scary part for Megan and her mom is there goes their brand-new laptop and if we can't find it (or it gets destroyed upon landing), they're out a whole computer. The scary part for me was that if there were any more software glitches, there was no way to fix them, no way to get any further data, and we would probably lose the whole craft. Everything hung on some sketchy assembled-at-the-last-minute free software. Yeah. Ouch.

Fortunately, that didn't happen. the data transmitted through the entire flight, and we got continual updates of images and GPS data every five minutes. However, we did have another glitch. It turns out that the balloon had a leak, and when it reached 766 feet, it hung there for a minute and then sank back to the ground. However, it did manage to travel about 30 miles (north instead of our predicted east) and it got so high that we couldn't see it anymore. As a bonus, as if to say "This was a really great try and you deserve a nice break", the balloon landed within a couple hundred feet of a major road near the next town over. But as if to say, "Wait wait, I'm going to throw one last challenge your way" it landed in a giant patch of blackberry brambles next to a big open field. It took us a half hour to fish it out:



We finally got it out after I hacked at the brambles for 10 minutes with a pocket knife, but not before the balloon popped on the brambles, preventing us from patching the leak and trying a third time. Still, it was a crazy fun project, and despite the fact that we only got about 2% of our desired height, we got some incredible pictures. Remember, all of these were transmitted to the ground from a laptop floating hundreds of feet in the air:







Not bad for less than $500 worth of parts and only two days to assemble them.

And if you think this is the end, you are sadly mistaken. Early on, my brother and I decided we wanted to try this too. He is a photographer, and he thinks he can get better pictures than the ones taken by 1337Arts. Our setup will be much closer to Project Icarus than Megan's was due to the fact that my brother wants to send up a professional camera, and we won't be able to transmit the images. However, if it goes well, I want to send up a spare laptop on a third venture and try Megan's experiment again. I would like Megan and her mom to come with us and help us out with these projects; it's only fair that after their hard work, they have their hands in some successful high-altitude photography. We have half of our parts already, and we are planning our first launch for spring.

Look out, 1337Arts, here we come.

Trip to Space
Progress: 6.47%  Flight Time: 0:09:42
Solar Array
Progress: 6.47%  Power: 65W

1 comment:

jmpf said...

great looking pictures and videos! when can we see the next launch? -- you should come check out the diyship forums for more information and an evolving community on DIY weather balloons