Posts tagged tracks
A slower pace and Astdroid’s own spinoff
Nov 4th
A couple days went by without hearing new of Astdroid, but I remained optomistic. A week went by and my hopes faded slightly. At about a week and a half in I admitted some amount of defeat and purchased a new phone. Now that it’s been a couple weeks I have no choice but to wonder if Astdroid will ever be found.
It’s been slightly demoralizing in some senses, the launch and communication with the phone had gone so well, but in another sense it wasn’t entirely unexpected. Sending a phone into space and recovering it is an ambitious goal to begin with. Couple that with bare bones equipment and a first try and it would have been very impressive indeed to have had a 100% successful inagural launch.
The silver lining for a failed first attempt is found in lots of lessons learned. In studying our data, reading of other lost balloon accounts, and taking a hard look at our code, I know our next launch will be a drastic improvement.
Perhaps, in some sense, things moved a bit too quickly. In an attempt to stay in the zeitgeist of the internet (and its notoriously short attention span) I wanted to get this project running and to its first success as soon as possible. Some things were overlooked, others consciously ignored in favor of a speedy schedule. I put alot of work in a short period of time, which was great, but with it came bugs.
After the failure to recover the first attempt at flying Astdroid the last thing I wanted to do was immediately sit down and start over again. I realized that doing so much so quickly had burnt me out.
About 2 weeks ago, having had some time to relax my brain, I revisited one piece of Astdroid that had bothered me and contributed to its recovery failure, the location tracking. Although the code I had written had worked as anticipated and we were able to watch Astdroid fly in real time across Denver, the battery it used to do so drained it incredibly quickly. In my tests it had gone from a full charge to 0% in about 2 1/2 -3 hours. As the balloon took much longer than anticipated to reach the height we expected, it’s quite likely that Astdroid wasn’t even on at burst, much less on the way down.
I have since completely rewritten the location tracking code, breaking it up into a separate project and greatly improving the battery life achieved while running it. In my estimation, the battery life has been increased on the order of about 3-5 times (I’ll be able to be more specific with some additional testing).
Since the end goal of Astdroid is to provide an application that others can download to replicate the project, I consider this a big achievement in redesign. We all know that 2 1/2-3 hours of batter life simply won’t due if the end goal is for this application to operate without the assistance of other devices for such an ambitious task.
Like I said, there were lots of lessons learned from the first flight. I’m approaching things this time around with a little more attention to detail and a lot slower pace. You’ll continue to see updates here, but they may be fewer and farther between. Rest assured though, this project is by no means dead.
In breaking out the live tracking code of the Astdroid project and taking a deep look at it I realized it had a use as a utility on its own. I decided to spawn off a project for that functionality called LiveTrax. I consider it a tangible byproduct of Astdroid’s software achievements.
LiveTrax is currently available in the Android Marketplace for free. With a couple button pushes your phone is tracked in real time and a link to a unique url is provided for your to share with your friends and family.
In addition to tracking a phone as it hitches a ride with a weather balloon I envision it to be useful for hiking, road trips, and races. Even something as simple as “I’m on my way to meet you, here’s where I’m at now u.livetrax.me/m/____” It’s often illustrated how many great products we use today originated from NASA ventures, I consider LiveTrax spawning from Astdroid much the same way 
I’ll be adding some cool features in the coming weeks so stay tuned! I’d love to know what you think!
MIA
Oct 4th
Last Saturday was the big day. After an incredible fundraising period, lots of work programming, designing, and testing, and hours spent on the phone making sure everything was ok with the FAA, it was time for Astdroid’s first launch.
We woke up at 5:30AM, drove up to Boulder, CO (about 40 minutes from Denver) rolled out our big tank of helium and got to work. My wife (@erinpier) streamed on her phone via UStream allowing people from all over the world to watch. The weather was perfect. We had donuts. It was great!
We were scheduled for a 7:30 launch time and were in the air at approximately 7:46. The balloon rose directly over head for about 30 minutes before moving south as we had anticipated. We noticed it rising a little slower than expected as we watched Astdroid Live while the HTC Evo and the Astdroid application were sending back data. Other than that things looked good. The application told us the craft’s position, temperature, speed, elevation, and battery percentage. It was working marvelously!
Once the balloon was less than a speck in the sky we decided to reward a launch well done with a coffee and a small breakfast before heading back to Denver and ultimately preparing for the recovery part of the journey. At about 9:30 we saw that Astdroid had stopped reporting it’s location at about 30,000 feet, just like we had expected, as the cell towers were no longer in range for a 3G connection. No worries, we still had updates from our SPOT GPS (though only at 10 minute intervals as opposed to 30 second intervals) and knew we were still on course, drifting across Denver heading for Colorado Springs.
At 10:11AM the updates from SPOT had stopped. We waited for a bit knowing that sometimes the SPOT missed an update. 20 minutes became 30 minutes became an hour. Something was wrong. At 10:11AM we went from everything going as expected to not knowing what had happened.
So, here is where things become speculation:
- The balloon rose more slowly than was expected in part due to a heaver payload than recommended. Our payload was just under 880 grams while the max was 905 and the recommended was 605. We figured that accounted for the slower rise up. We also believe that means the balloon never reached its ceiling (90,000 feet).
- Because of this, the balloon likely popped at a lower altitude. My guess would be somewhere in the neighborhood of 60,000 feet. Though that’s only a guess.
- The transmission at 10:11AM was the last one before Astdroid reached ground, landing in a way that didn’t allow for it to communicate its location. Upside down? In a valley? In pieces? Who knows.
We went and checked out the last known location. No luck spotting Astdroid. The unfortunate thing about the SPOT GPS is that only updates every 10 minutes. If that was indeed the last update before it hit ground it might have traveled an additional 30 seconds… it might have traveled the full 10 minutes… that’s alot of range.
It was beginning to get dark so we went home that evening empty handed.
That night we took a hard look at the data. Here’s what we came up with.
There were 2 possible scenarios. The SPOT GPS had stopped operating for an unknown reason in midflight, thus leaving us with absolutely no idea where Astdroid may have landed OR we were correct in assuming that it had stopped transmitting due to it’s position on the ground. Being that the second option is the only one that would allow us to recover Astdroid, we pursued that avenue.
Here’s what we have from the SPOT GPS tracker:
Using that we were able to take a look at the speed between updates:
The distance deltas dropping from 9 miles to 4 near the 3rd and 2nd to last points lead us to believe the balloon burst and fell before drifting to the ground.
As you can see this aligns nicely with the idea that the balloon burst and fell (quickly) as expected before the parachute could open (at an elevation of 60,00 feet there would not be enough air for the parachute to open) before drifting.
This would mean that if the payload were moving consistently with these speeds, or better still, slowing down, then the payload would have been travelling at about 20 mph. At 20 mph with a maximum interval of 10 minutes that would mean that Astdroid were within a 3 mile radius of the last known point. Adding to that the knowledge that Astdroid was travelling almost due south at it’s last update, we conjectured that it fell within a cone of about 3.3 miles. So that’s what we decided to explore.
The next day we decided on another recovery attempt, armed with much better knowledge. We were able to do 1 mile sweeps from the last known point, but we saw nothing.
Blue line - The SPOT GPS Track. Orange polygon - 2 mile cone. Yellow polygon - 4 mile cone. Red line - GPS track of explored area
We covered almost 4 miles on foot with no signs. To cover the entire area is almost impossible as it grows exponentially with each mile in radius. The area is populated, but not heavily.
I’m still optimistic, hoping any day somebody will call and say they found a strange cooler in their backyard. I have some ideas about enlisting the help of some local geocachers to go explore. However, it must be said that there’s a possibility it will never be found.
There were many, many lessons learned from the experience and this has by no means ended the project. Next time we will be better armed with knowledge, experience, and data!
I’m posting all of the data from the Astdroid launch here. If you’ve got an ideas, send then my way!
Google Earth KML of SPOT GPS track, Projected Cones, and Explored Tracks
I’d love to have written this post sharing incredible photos of space with you, but this time it was not to be. Next time, next time we will have our due success.
