Blogs

I've been a big fan of the Parrot AR.Drone since it was first released in early 2010. Although it's not a real drone as we define it (capable of autonomy) since it doesn't have GPS or a navigation function, it is a remarkable piece of engineering and I think of it as the "gateway drug" to real UAVs. I've heard that more than a 100,000 have been sold, and many of those users will catch the bug and want more, bringing them to communities likes this one and the more sophisticated UAVs discussed here. 

Parrot has been kind enough to send me review units of both the original AR.Drone and now the new AR.Drone 2.0, which was released last month ($299). The 2.0 version adds:

• Much improved video (it's now very clear with its new HD camera and almost lag-free)
• A magnetometer for what we call "simple mode" flying, where you can ignore the orientation of the quadcopter and just fly relative to the compass (so tilting your iPhone/iPad down will fly it north, for example, regardless of what orientation it's in). 
• A pressure sensor so it can control altitude above the 5m range of its sonar sensor.

Indoors, the main advantage is the better camera and the pressure sensor means that it doesn't bounce up when it flies over furniture, the way the sonar-only AR.Drone 1 did. 

Outdoors, the advantages are more obvious. I flew it last weekend with the Bay Area crew (see post here) in some pretty stiff wind and we were all very impressed. Here are some quick observations:

• You can now change the settings to allow altitude of up to about 100-120 feet. Parrot says the altitude is unlimited, but in practice I wasn't able to get it any higher than that (perhaps because it won't go further than its wifi range).
• The main limiter on outdoor use in any direction is the wifi range. I was able to get it to fly about 100-150 feet away but then it lost the wifi connection. The good news is that when that happens, it just hovers in position until the connection is regained. So you just have to walk closer to it until it gives you control back. 
• Loiter (which uses the optical flow camera underneath the copter) was very good. Close to the ground it's really locked in (the more unique objects underneath it that it can see, the better; regular grass and dirt doesn't offer many unique features, so it will drift a bit over them), and the higher you go the less effective it is. But even at 100 ft in a stiff wind, it will only slowly drift from its loiter position, and it will stay within a few tens of meters for as much as a minute. 
• Wind-handling is as good as you can expect for a relatively small frame and low-powered motors. It keeps its stability in the wind well, but does require some pretty aggressive tilting to not be blown downwind. 

Overall, I think the 2.0 version is a significant improvement. The new HD camera is the biggest win and the indoor altitude hold performance is much improved by the baro sensor. Outdoors, I think the advantages of the higher altitude option are not as great as they could be, due to the Wifi range limits, but it does widen its performance range somewhat. 

For people who don't already have a Parrot AR.Drone, the improvements in 2.0 definitely make it more attractive. If you've been waiting to get one, now's the time. But if you've already got a 1.0, you may not find them quite enough to upgrade. 

As before, the AR.Drone is really optimized for indoor use, and it's hard to find a better platform for that, from stability to reliability and safety (relatively soft props and a foam protective ring). I think you'll see more and more people using the basic frame for indoor automomy research with replacement controller boards.

Bottom line: it really is a beautiful bit of engineering and design and an inspiration to us all in terms of ease-of-use.

UPDATE: here's a video of someone flying 300ft away. They must have a cleaner wifi environment than I did!