Building a UAV for photo mapping - Autopilot



Ahh!! CRASH and BURN!!

I had planned to use the Pico-Pilot and Pico-GPS for the autopilot in my UAV, but I have now discovered that since Jan 2007, they have been classed as MILITARY technology and are controlled by US Export License regulations. Specifically the regulations cover,


a. “UAVs” having any of the following:

a.1. An autonomous flight control and navigation capability (e.g., an autopilot with an Inertial

Navigation System); or

a.2. Capability of controlled flight out of the direct visual range involving a human operator

(e.g., televisual remote control).

b. Associated systems, equipment and components as follows:

b.1. Equipment specially designed for remotely controlling the “UAVs” controlled by 9A012.a.;

b.2. Guidance or control systems, other than those controlled in Category 7, specially designed for

integration into “UAVs” controlled by 9A012.a.;

b.3. Equipment and components specially designed to convert a manned “aircraft” to a “UAV”

controlled by 9A012.a.

Note: 9A012 does not control model aircraft.

Despite the last sentence, UNAV, who make the Pico Pilot have now told me that none of the applications for export licenses they have made this year have yet been granted. Back to the drawing board.

The other common low cost option for an autopilot seems to be based on the FMA Co-Pilot for flight stability with an additional board such as the RCAP2 plus a GPS receiver for navigation. While a cheaper alternative, I had already discounted this approach because it is based on thermopile sensors. For my terrain, I cannot get a clear 360degree view of the horizon to calibrate the system before launch. In addition, the various different terrain types , forest, grassland, lakes etc. could give problems in flight, irrespective of the temperature differences that can occur if different parts of a valley are in sunlight or shade.

During my initial research into autopilots, I also looked at the Paparazzi project. While there is a wealth of open source stuff there, the current Tiny autopilot still uses thermopile sensors for stability, although it does have an on board GPS unit for navigation. An all singing, dancing IMU with gyro's , mangetometers etc. is under development. Although all the designs are published, there is still no commercial source of assembled units or PCB's.

A recent post on this forum (can't find it now), talked about the the UAV development board from Sparkfun. I had a brief look at this, but initially discounted as they claim that the firmware is a guideline only. It is also written in assembly code and I am far to old to start writing in assembler again. Still I shall have another look at this over Christmas, as the board does have a proper IMU with 2 gyros and a 3 axis accelerometer.

Conclusions and Questions

  1. I cannot use my prefered autopilot option owing to US export regulations.
  2. What other, non US manufactured, commercial autopilots are people using?
  3. Any other suggestions for a home built unit with an IMU rather than thermopiles?


Building a UAV for phot mapping - Previous Posts

  1. The Goal
  2. Camera Selection
  3. Packing it all in
  4. Autopilot

Views: 3323


3D Robotics
Comment by Chris Anderson on December 23, 2007 at 11:35am
Paul,

That sucks, and I'm once again embarrassed by my government. If you'd like we can talk offline about ways to get around this. No harm, no foul, etc.... Use the private messaging system (envelope icon above) to ping me if you'd like to discuss.
Comment by Jack Crossfire on December 23, 2007 at 6:19pm
Sounds like more ways to lock out the startups while the Lockheeds & Boeings get a free ride.
Comment by Patrick Egan on December 23, 2007 at 7:08pm
There is always micro pilot...
Comment by paul hubner on December 24, 2007 at 9:39am
..and there's always Ebay. Or the RC groups forums for personal purchases. Or a friend in the US who would send it to you, etc,etc.

I am too am embarrassed by the outmoded laws that only keep the honest honest!

Paul
Comment by hotelzululima on December 25, 2007 at 8:59pm
Uh, as a follower of the paparazzi project, I think you may be a little confused as to the actual capabilities of thermopile sensors, they were selected NOT for cost but for accuracy over IMU approaches instead.

Except for very certain conditions(ie forest fires etc) they are as accurate or more accurate than IMUs(gyro,mag,accel) in the equivalent price range or even far above.

thought you might like to know


gwen - researcher
Comment by Paul Gregory on December 26, 2007 at 12:52am
Hi Gwen

Having not yet tried any form of autopiot, my comments were not based on any practical experience - just trying to think through the issues. Certainly in the FMA instructions in the calibration section, it states.

" Measure the temperature difference between earth and sky.
a. Take the model to a spot (grass is best) representative of the area where you will be flying. The aircraft should be atleast 100 feet (30 meters) from anything that radiates heat (heat from buildings or parking lots can affect calibration)."

If you look at my launch/landing area photo in the first post , you can see I have the house to one side and trees on the other - both closer than 30m. This combined with the steep hills and valleys and the great variety of terrain types led me to think that I would have problems with any system that uses thermopiles.

If, based on your experience, you are saying that these factors are not important, then that is very useful to know. I have also not seen any mention of them being more accurate than IMU systems - do you have a reference?

Paul

3D Robotics
Comment by Chris Anderson on December 26, 2007 at 8:48am
Paul,

We use the FMA thermopile on GeoCrawler 1-3 and have no problems with heat-radiating areas at close range. That's because we only engage the autopilot after we're at an altitude of 100m or so. I assume you'll be doing the same thing--taking off and landing manually, with the autopilot disengaged--so I don't think you have to worry about this.

That said, the Paparazzi hardware/software is technically challenging to get right, so the Picopilot is by far the easier path.

T3
Comment by Krzysztof Bosak on August 29, 2008 at 5:00pm
"{we}...have no problems with heat-radiating areas at close range. That's because we only engage the autopilot after we're at an altitude of 100m or so."
And that's the point. Thermopiles are working good if you provide a biological pilot to the autopilot to land/takeoff the whole thing. This diminishes the number of potential users by few orders of magnitiude. The problem is, you really need a MSc in applied maths/physics/related to make a good IMU at home.

Comment

You need to be a member of DIY Drones to add comments!

Join DIY Drones

Groups

Season Two of the Trust Time Trial (T3) Contest 
A list of all T3 contests is here. The current round, the Vertical Horizontal one, is here

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service