GPS-guided Rocket Recovery Project

This is the home page for my GPS-guided model rocket recovery project (last update September 2011).

The idea is to have the the rocket fly itself back to the launch pad using a steerable parachute and onboard GPS.

At this point I have the “Rev 4” PCB hardware and software working well. The navigation has worked a little bit on a couple of of test flights, but I’ve had problems with the parachute lines tangling and twisting – I got side-tracked for a whole year using a baseline for steering the parachute that was far too small – this was the main cause of the twisting. I’m making changes to solve those problems.

All the hardware schematics and almost all the software source code are posted here. Here is a quick rundown of the hardware specs:

  • CPU: Microchip PIC32 MCU (PIC32MX440, with the USB interface)
  • GPS: GlobalTop PA6B (updates at up to 10 Hz)
  • Telemetry: Microchip MRF24J40MB 100 mW IEEE 802.15.4 radio transceiver (2-way)
  • Altimeter: Freescale MP3H6115A pressure sensor (launch detect, electronic parachute ejection, beeps apogee altitude)
  • Hobby servo outputs: Up to 2 servos (to steer the parachute)
  • Power supply: 2 x LiIon cells, Microchip TC105 switching DC-DC converter
  • Outputs: 4 high-current outputs for parachute ejection and staging
  • Camera: Controls a still or video camera in-flight
  • Flight log: Stored in flash memory (lat/lon, speed, altitude, status…)
  • Ejection detection: Via CDS cell (light sensor)
  • 3 push-buttons, 3 LEDs
  • Piezo speaker output

Here is a quick guide to the project info on this site – see the “Rockets” category for other rocketry-related stuff:

  • Project intro from 2008 (aka Nevada Dave and the Rocket-Eating Trees). Includes slides describing my original ideas and goals for the project, and progress thru the first couple of hardware versions.
  • Flight test notes (pdf). Notes of every test flight from the beginning (mainly for my own reference). There’s lots of technical nitty-gritty here, but it has the most up-to-date info (from time to time I post news here on how things are going, but these notes get updated after every flight).

Similar & related projects

  • Pat Horton’s SPRS – I don’t know him, but Pat built & flew a working radio-controlled (not autonomous) “Steerable Parachute Recovery System” back in 2001. It was finding his web page that convinced me this project was possible, and got me interested in the NASA NPW5 parachute design that I’m still using.
  • Mike Passaretti – Mike’s goal is similar to mine – fully automated recovery using a steered parachute and autonomous GPS navigation. Mike seems like a very experienced engineer, and I think has a good approach, tho different in some ways from mine. He started his project in 2011; we’ve exchanged a few emails and swapped notes, and may end up collaborating or something. Have a look at his site.

If there are other people who should be listed here, please let me know and I’ll add them. I know similar systems have been working in the military for a long time, but I think the hobbyist community is just getting there now.

10 thoughts on “GPS-guided Rocket Recovery Project

  1. Dave,
    I’ve just become aware of your “Fly it back to the pad” rocket project. This is,Hands down,the most ambitious,exciting,and inspiring project I’ve ever heard of! Thanks for pumping up this mere mortal rocketeer and for sharing your progress on this awesome difficult endevour!
    STEVE MAHONEY

  2. GREAT project! Love it! What a fabulous idea!
    After reading some of your flight logs, it seems that two major problems that you are encountering are reliable data in the pressure sensor and also the GPS unit (stemming from being able to acquire the satellites). This is just a thought and suggestion from waaaaaaay left field, but have you considered using a 3-axis accelerometer? Theoretically, you could track the flight in 3d w\o any interference (pressure sensor being affected by engine, GPS needing relatively stable position to get sat fix, etc…). You would then calculate the trajectory back to the original point. I dunno, just a thought! GREAT project!

  3. Hi Jon,

    Thanks for your comments! Only the GPS is used for navigation; the pressure sensor (altimeter) is used for launch detection and apogee detection (to do parachute ejection). Altho it’s a bit noisy (I’m working on it), it’s plenty good enough for that.

    And the GPS has been in fact pretty reliable. I think a larger problem might be that the whole electronics bay (including the GPS) is swinging back-and-forth under the parachute lines, which confuses the calculation of the direction of flight. To try to deal with that, my friend Boris and I are building a rocket glider that will use the same electronics to steer the glider – we’ll see how that works later this year.

    I have thought about adding extra sensors (like a magnetic compass or a 3D accelerometer), but I don’t think they really add much to the navigation and of course they take space and make things more expensive. I suppose if the swinging problem turns out to be severe, I could try adding an accelerometer and use that data to “subtract out” the swinging. If all else fails, I’ll try it. (Thanks for the suggestion!)

    As I write this, I’m getting ready for a couple of launches – one at CMASS on Saturday 16 July, and the OROC Desert Heat launch in Brothers, OR the following weekend. I have updated software (more robust and more logging of internal events) for those. I also have rigged up “808” keychain cameras to video the parachute in flight – maybe I can learn something from that.

    I got in 3 flights at CMASS on 30 April 2011 (no onboard video then), haven’t had time update the blog with the results of those yet. After I get back from Oregon, I plan to update things here with the results from that and the two upcoming launches.

    Cheers,

    –Dave

  4. Watching the videos, I see what you mean about the swinging. There’s the component of the entire mass (parachute + electronics), then there’s swinging electronics bay underneath it… Not to mention the entire mass may be tumbling head over heel! Hmmm… Well, it probably wouln’t be as much fun if it WASN’T a challenge! =) Looking forward to seeing the results of this years launches.

  5. I know the videos make it look like the thing is “tumbling head over heel”, but I think that’s an optical illusion – the rocket body (and electronics) is always below the parachute, but when it’s swinging underneath, and video is taken from directly below, it looks like the rocket is on top. But it’s not.

  6. Dave,

    I have used a design that mounts the GPS and camera at the parawings pivot point to solve both problems.

    Postage stamp GPS / Camera have only made that easier.

    e-mail me if you need more details.

  7. Hi Tom,

    I am _very_ interested. (And sorry for my slow response – I was in Oregon at the Desert Heat launch…)

    You mount the electronics directly at the end of the parachute steering lines? Have you had success with automatic GPS navigation of the parachute/rocket?

    I looked at your website – not much there.

    Can you tell me more?

  8. i am interested in this project those it work?
    how can i do it? do you sell the gear to do it?

  9. How is the project going?

    I’m in the early stages of (hoping) designing and building
    an RC-steerable parachute. Initially attempting GPS
    steering and autonomy seemed overly optimistic. I’d just
    like to bring ‘er home to avoid a lot of hiking. The RC
    part seems like it would add fun to the equation.

Leave a Reply

Your email address will not be published. Required fields are marked *