Hacker News new | ask | show | jobs
by uptown 4480 days ago
The Phantom 2 are pretty great drones ... but they're also prone to "fly away" where they decide to take-off on their own - ignoring your commands. There's a bunch of videos on YouTube showing people's Phantoms flying away. Like this one:

http://www.youtube.com/watch?v=HkQ9eB7M7iQ

This video purports to help prevent that from happening, though it's not guaranteed to prevent your $1000 drone from making a break for it.

http://www.youtube.com/watch?v=2bxjL7wFyb8

5 comments

Watching an uncontrolled drone strafe an overpass at street level and split power lines for a commuter train, really increases my sympathy toward those who want to regulate these.
We really should also regulate pigeons.
Pigeons usually do not have fast-spinning blades attached to them. I am not really for regulation, but these things can actually hurt people [1]. RC Helicopters (with one large, main rotor) have actually killed people [2].

Hobbyists must take great care to operate their drones safely, and companies should assume some liability for runaway situations like this if they ever result in injuries.

Warning, pictures of injuries:

[1a] https://www.google.com/search?q=quadrotor+injury&tbm=isch

[1b] https://www.google.com/search?q=multirotor+injury&tbm=isch

[2] RC helicopter fatalities: http://www.dailymail.co.uk/news/article-2413231/Roman-Piroze...

So can people with knives or swords, or baseballs propelled by a bat. Anyway, I suspect that all of those quadcopter injuries occurred while the owner was setting up the quadcopter. Every story I've read (mostly from my frequent visits to RC hobby forums) is from people attempting to test functionality on a bench with the propellers on, or perhaps a few people who attempt to take off with their aircraft right at their feet. The RC helicopter death is a valid point, but that's the only RC aircraft related death I know of, and helis like his pretty much self-regulate because they're expensive, require mechanical knowledge to setup and maintain, and will mostly likely never get off the ground if flown by someone who doesn't know what they're doing.
>Pigeons usually do not have fast-spinning blades attached to them

BRB attaching circular saw to pigeon

Sure, everything has to be regulated; but how do you propose to regulate and then how do you propose to enforce those regulations?
I'll leave the specifics to the better informed, but certified 'kill-switches' and 'no-fly-zones' seem like obvious places to start.
DJI is massive player in the commercial drone space, so when regulations get written, you can bet DJI will be one of the corporations writing them.
That was rather spectacular. I couldn't help but think that someone else controlled it, it was so many times close to run into power lines. Are there any believable theories as to why that happens?
Drones are getting pretty incredible. This one is almost indestructible: https://www.youtube.com/watch?v=pl2Z9N4Q82g
It fails as soon as one of the rotors deforms. That doesn't make it very reliable.

I wonder how many rotors they went trough the entire lifespan.

I backed their Kickstarter, and have also met the guys and seen those airframes in person at their "Game of Drones: Flight Club" meetups in SF [0]. You're right that they still use normal propellers, and that those will be the weak spot when ramming into things. But the point is simply that the airframe itself is sturdy and nearly indestructible. They're designed to battle in the air against one another, which inevitably causes lots of falls and prop replacements. They're not meant to be "reliable" in the sense that they could fly autonomous missions repeatedly or serve other commercial purposes. They're meant to be beaters.

[0] http://www.gameofdrones.biz/#!sumostory/c2kw

Have they ever thought of using carbon fibre props, you can 3D print CF now
CF props are quite common for multirotors, but I don't think they'd be good for battling simply because they're significantly more expensive than plastic props, and for battling you're not really concerned with efficiency or prop deflection.
I've had a Phantom for a while, and more recently the Phantom 2. Have never had any issues whatsoever. I don't discount that these happen, but as with most things, there's a pretty big sample bias in complaints.

Also of note, winds aloft are often not the same as winds on the ground. Given top speed on these guys is around ~20mph, any wind above that and you've got a fly-away. Again, definitely doesn't account for all fly-aways, but I would venture that it does explain some.

Is the controller link digital or analog? If it's digital why don't they checksum and/or sign packets to avoid continuous interference doing anything but triggering failsafe mode?