A project by Cambridge University Spaceflight
Project Squirrel aims to explore the feasibility of using an Android phone as a multi-purpose flight computer for high altitude balloon launches. A first proof-of-concept launch took place in June 2010, with a further test taking place in May 2011.
The second major launch took place on June 23rd 2011 using a Nexus One phone, with the balloon reaching 36.2km altitude (a new UK altitude record). The Nexus One was shown to work successfully as an entirely standalone flight computer, with four custom apps running to perform the required functions. On returning to the ground by parachute, a precise GPS location was received by text message, allowing the payload to be located with ease.
Many photos and videos were captured (up to an altitude of 25km), and sensor data was logged for the duration of the flight.
Takes photos and records video according to a schedule (for example, 12 photos with a 20 second interval, then 1 minute of video, and repeats), storing them to the SD card.
The underlying Android CameraService is prone to occasionally crashing in the event of low memory, especially when recording high resolution video (several bugs have been identified in the MediaRecorder framework), and as such extensive measures have been put in place to ensure that any errors encountered are handled intelligently, without affecting the functionality of any of the other apps (as far as possible).
Squirrel Log takes advantage of the on-board sensors present on most Android devices, as well as several other data sources, to produce a useful log of the flight. This includes:
A simple text message based phone finding app. It replies to text messages with any available GPS data, or failing that, with its location as determined from the mobile network.
This app operates in conjunction with a Radiometrix NTX2 transmitter connected directly to the headphone port of the phone, and allows the phone to communicate with the ground when it has no mobile network connection (anywhere above a couple of kilometres). It generates RTTY tones to transmit the payload telemetry to the network of ground-based listeners, using the integrated GPS. Part of the RTTY string is encrypted, as a simple proof-of-concept for a cryptographically secure channel of communication.
An example RTTY transmission: (telemetry begins after 6 seconds)
Additionally, live images are transmitted using the SSTV (slow-scan television) protocol - a 320x256 image takes roughly 2 minutes to transmit.
An example SSTV transmission (decoded image here):
The Nexus One was housed in a polystyrene box (with a hole for the camera) which could then be easily suspended below the helium balloon. A simple quarter-wave dipole antenna was built into the bottom of the box, to which the NTX2 transmitter module was connected. The box was covered with insulating reflective material.
The Squirrel payload launched from Churchill College, Cambridge (the CUSF launch site) at 13:15 GMT on Thursday 23rd June 2011, suspended below a 2000g Hwoyee meteorological balloon. Also attached was a separate payload with a u-blox 5 GPS and transmitter. Despite heavy rain and windy conditions, a short window of good weather allowed us to fill and launch the balloon.
Due to an unforseen frequency clash between the Squirrel transmitter and the other payload on the same balloon, it was deciced to turn off the Nexus transmissions for this flight since it had already been tested with great success on Squirrel 1.5 (on that occassion it saved the payload when the main flight computer lost GPS lock).
The payload was tracked for nearly 5 hours by a number of listeners from CUSF and the UK High Altitude Society, and reached an altitude of 36.2km above sea level - a new UK amateur altitude record. At this point, the balloon burst due to expansion at low pressure, and the payloads returned to earth under the pre-deployed parachute. The landing site was a tall tree in a field in Exning, Newmarket (Google Maps), and the Nexus One sent us a text message upon landing with it's exact GPS coordinates.
Recovery was made difficult by the height of the payloads in the tree (roughly 20-25 metres), and attempts for retrieving the two boxes were abandoned until after the summer university vacation. On 10th October, a return trip to the field with the help of a tree surgeon finally extracted the payloads - even after 3.5 months in a tree the Nexus One was in perfect condition, and the SD card contained all of the photos/videos and recorded data.
The phone successfully recorded photos and videos up to an altitude of 25km, at which point the Android MediaRecorder framework crashed (a problem we had encountered before and had made suitable precautions in the app in the event that it occured during flight).
All of the photos and videos can be found on the Squirrel 2.0 Flickr Set, or in the embedded slideshow below.
Below are four graphs showing some of the data collected by the phone's sensors. The Android GPS was limited to 18km altitude (see COCOM Limits), but GPS data for higher altitudes was logged by the other payload (which used the u-blox 5 chip which isn't altitude capped in the same way).
Through the work of Squirrel 1, 1.5 and now 2.0, we have been able to show that an Android smartphone can fully replace a conventional flight computer for high altitude balloon flights, operating with external temperatures as low as -60°C and at altitudes in excess of 30km. The convenience of having a readily available consumer device which includes multiple sensors, a camera, GPS and SMS functionality puts this sort of experiment within the realms of any science enthusiast.
Furthermore, it seems likely that with the increasing prevalence of low-cost, mass-produced smartphones, experimentation and testing such as we have demonstrated through this project will become increasingly attractive for not just high altitude research, but the technology industry as a whole.
To find out more about this project, please contact Edward Cunningham, firstname.lastname@example.org.