TaigaCam, ROBUST plane for aerial photography

Don't do this at home with other planes!  ;-)

In the first episode motor, rudder and tail servos are not yet installed.

Plane is home made, there is no ready kit available.

- Wingspan 150 cm
- Length 120 cm
- Empty weight without batteries 1.3 kg
- Maximum load with batteries is about 1.2 kg
- Maximum takeoff weight is about 2.5 kg
- Flight speed is about 50 km/h, maximum about 80 km/h
- Flight travel with one 2100 mAh battery is about 7 km, flight time is approximately 8 minutes.
- In plane there is space for 4 batteries, so flight time is more than 30 minutes
- Space for autopilot, batteries and payload is about 30 x 7 x 5 cm.
- Wing is made from EPP foam, it is made with CNC hot wire cutter
- Fuselage is made from EPP foam and PVC pipe or fishing rod
- Price of the plane frame without servos and motor is less than 100 euros
- Price of the plane with good servos, motor and wires, but without batteries and remote control is about 395 euros in Finland, in other countries much, much less...
- The building time is about 8 hours

Plane homepage in Finnish: http://sites.google.com/site/taigacam/

E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones


  • T3

    "EasyStar and Ardupilot for light and short flights (400USD)"

    How did you come to that price ;-) It is not possible to start intensive RC hobby with that much money.

  • TaigaCam is not very streamlined. Wing surface is not very smooth, it is cut with hot wire cutter. My camera stabilizer test systems are also quite large.


    In battery test I used 4 year old ThunderPower-battery. Then I tape Garmin GPS over the fuselage. I flight manually as long as it was possible and then landed. I downloaded log file from GPS, and look there start time, landing time, and travel between these. There I got those values (7 kilometers and 8 minutes).

    In Cropcam I use 4 ThunderPower batteries, and I use always 2 cameras. Safe flight time is about 30 minutes (=30 km) and with parachute 25 minutes (=25 km).

    PS. I just came from test flights. First start was disaster, and plane crashed to ground. Two aileron hinges broke. I repaired these, and after one hour from the crash I made three more flights. There seems to be still some adjusting in Ardupilot gains...

  • I think there is still niche for all these planes:

    - ArduCopter, if you want to take photos nearby (800 USD)

    - EasyStar and Ardupilot for light and short flights (400USD)

    - TaigaCam and Ardupilot, if you want to build yourself a robust plane, and want more heavy payload (600USD)

    - Cropcam, if you want to take photos from prairie fields or other open areas (7000 USD)

    - Pteryx, if you want to have more robust plane for large areas (8900 EUR)

    - Bramor, Gatewing X-100, Mavinci Sirius, etc, if you have more money (>20000 USD)

    - Those big UAV:s that can fly thousands of kilometers to Atlantic, and measure storms etc. (Millions dollars)


  • T3
    I have calculated a few things and got a little surprised. At half weight of Pteryx the plane consumes slightly more than twice the amperage (assuming lipo 3S). Not sure how are you going to cover grounds with that, when you simply add batteries it will be a brick, but a very hungry one (and no longer that resistant).
  • sweet thing.... panzer plane ;)
  • That looks like a perfectly good landing site to me :-)
  • I like it !!
  • T3

    Dear Alpo, I am not hiding, your operating conditions appear next to impossible.

    Let's put together things:

    -you can land on a stump, it is like landing on minefield

    -the payload is fragile, as it contains camera

    -the landing place is small field maybe 200x200m

    -obstacles aroudn are high, a loss is a total loss

    -modem reception throught trees is weak, 2.4GHz systems might be unaccaptable

    -Crucial: the areas taken in forestry are high. This VTOL flying very difficult, also the obscurity of video transmission through trees limits FPV, the endurances of quadcopters are always limited to winged machines

    -you have a law limit of 1.5km VLOS


    IMO you cannot operate anything reliably in those conditions, but you could if you get rid of the last point and chose small landing place without stumps. Even UAVs in mil zones break if they simply land on the rocks or stumps.


    On other hand, taking cropcam as reference is misleading: the sturdiness of hobby grades ara a joke. The stiffness of fuselage walls of pteryx feels like 3 times more than any flying model, the resin is the same by name (polyester), but it is not that brittle because of quality. In fact we hang the plane on the wall by the tail like a trophy hunted animal.

    Cropcam is known for extremely cramped space inviting crazy cabling, insufficient space for foam around batteries etc.  I believe the worst competently designed UAV from composites would be more resistant than cheap hobby sailplane fitted with autopilot. I believe Easystar is one option because you can afford spare wings and because you can cyano-glue it often in place - but the range is disappointing.

    Part of cropcam mayhem is the amount of servos and control surfaces. Let's make it clear, guys got funding for the autopilot business, made it well and their standard testing plane was promoted as photomapping product at some stage. None of the operational issues have been solved since putting it on the market, because it was rather a reasonnably priced demo plane.


    Now let's take our worst crash, Pteryx prototype 1, less rigid fuselage than the rest. Only maybe 300-400 takeoffs but al lof them for research purpose (changing load, COG, firmware). 1 tree hit (huge operator mistake), one spin to the ground from 50m (winch takeoff ok, but the ballast moved upon takeoff).



    -broke 5mm balsa sheet horisontal stabiliser (somehow), you just put any balsa sheet and replace it

    -the wing is a total loss, but is less than 5% of the total price (it requires some real hand work to make it)

    -the camera head went ballistic like a missile and flew a parabola to the ground from 12m. Nylon camera mounting screw gave up, the impact cushioned tha camera is still in use and makes our next showcases (S90). The nose of the nose got traces of hitting a small rock waiting for it among the grass. Nothign worth documenting, gelcoat came off on 5x10mm area on the nose.

    -we spent afternoon to take the fuselage down. Ended with a traces of wing fallign like leaves all around, the fuselage finally felt fown like a bomb and hit a grass.

    Now the bad  news: no scratches. It flew next day (but we had the wings). It is annoying because the P1 required refurbishment in order to match the rest of the fleet, but since it works, we leave it as noob's testing plane.


    I do not believe you can fly Pteryx among stumps and have commercial profitability in forestry (but you can if you fly for construction site mapping). But half of those problems is that you started doing your own tinkering on a plane that in fact was not ready for forest operation. Crashed because of... Servo short? Cmon. We have three power zones, autopilot, low power servos, high power servos. Basically 60AMP ESC powers a motor that makes 30A peak duing takeoff, but some 30% of it at cruise. And so it goes. All those time wasteing issues you had are because the company started with cheap Elektra Pro model, it has achieved its limits of capability beyond demo plane.


    A treehit: we just said in a chorus: now that's a writeoff



    We got it, somehow the camera works, the nose cone has only minor scratch. We got lucky, but before we got this on the ground, it will be done anyway, we believed. Guess what...


    The outcome for the electronics: servo output connectors straightened with tweezers. The battery hit frontal wall and still operational. We could say we were lucky, but we werent. The plane hit so hard that the parts of the wing dissipated styrofoam core into the cloud of 3cm parts. Nylon screws gave up.


    Second case: the damage from the spin, repaired in the evening with low grade cloth 4x4cm and epoxy resin:



    And that's it. No more breaks. 4h ugly resin workmanship, 2h verifications, a few drops of glue, 2USD in nylons crews and a few precious rubber bands.


    I have sat on kevlar camhead. It survived, but is ugly. It heeps shape, the gelcoat is extremely scratch resistant but not as resistant as kevlar nor good fiberglass. Yet we are not maiking our fuselages from lowest grade cloth plus a lot of white paint, like RC model companies do. Our construction element IS the fiberglass itself, not the putty and paint.



    On the other hand, what I must reiterate, if you HAVE to takeoff for this place, it is as hard as landing on guided missile cruiser, therefore expect a crash rate not less than Orbiter UAV and prices not less than. Even Raven won't guarantee flying multiple misisons form there. Therefore the entire battle plan is a little flawed, there must be some permission to takeoff further from target or the final prices will be military on the long run. With Taigacam, I believe, you will save the platform but you will be beaten with all the rest requiring analysis.


    An example: once upon a time somebody disconnected the parachute servo. If was not used, but the 'High Commision for Pteryx safety' oreder changing the autopilto in such a way, it makes small audible moves with the servo just after booting. Now you can do it in open source projet, but once you fix that, there is reason why you wouldn't be beated by another detail that has been introduced by somebody trying to make a crazy loitering method using flying wing, whaterver. There are zillions of details like that, and it is unlikely that anybody tweaking existing RC model with autopilot will come close, therefore the syndrome of 'ever-repairing' airplane will persist.

    There were plans of making Pteryx from the foam, but the bottom line was that foam with high wing loading required for good turbulence damping and wind penetration is very fragile, and high endurance was unachievable

    (we have some examples from 2x2km mapping missions 70min each plus ample reserve - we use legal limits to the max).


    So please, do not take conclusions after using UAV from manufacturers that didn't took ANY effort on improving platform reliability. In our case we have had a few cable burnouts, servo jams etc all things we recover and when we see merely a close call, we reengineer the autopilot and a platform - other aviation manufacturers do the same.

  • If Cropcam or EasyStar (or probably also Pteryx) lands to a bad place, and one wing hits first to obstacle (tree, rock, stump), there will be several damages: Wing will break less or more. Then wing will violently rotate fuselage, and it will tear in front and in back of the wing fixing points. Probably also fuselage will snap near the tail.

    Here is Cropcam example. Parachute servo short circuits right after start(!), voltages went down, and plane fall  near starting place. This needs several days repairing. These high tech composites are not easy to repair. Also spare parts are quite expensive.


    Landings: I use automatic landing with Cropcam. Only if something seems to go wrong, I take manual control. (And always too late ;-)

    I have tested TaigaCam only with Ardupilot and manual landings.


  • T3

    Alpo, your requirements and cropcam issues were always clear to me, we have added a parachute and use industrail strength fuselage in Pteryx what is much better than any cheap RC plane. We did high T-tail and high wing, no steering surfaces in the wings just for that reason. Yes you will have problems if you have stumps all along, but cropcam fuselage is not a reference of sturdiness.

    However your Taiga Cam is so much sturdy I don't believe you could achieve something significant that you couldn't get with EasyStar, while BOTH lack some endurance - my point is that the Taiga Cam outlast its payload, whet means you will pay for that with some weigh and endurance. What is the surface area and pixel rsolution for forestry, if it is 1x1km it means something slightly larger than EasyStar would do the job.

    BTW are you taking off and landing in RC mode?

This reply was deleted.