Hacker News new | ask | show | jobs
by nlh 4482 days ago
This is supremely cool.

Tangential stupid question:

Obviously there's been a ton of buzz lately (and innovation) in the drone space.

What's the big technological driver that's allowed / caused this to happen? I don't know the details of the internals but it seems that the tech has been around for a while (radio controlled planes, servos, helicopters, smallish cameras, etc.).

6 comments

"What's the big technological driver that's allowed / caused this to happen?"

As with most things, several things had to come together in order for this to happen. A probably incomplete list.

1) Lithium chemistry batteries - unlike Nickel, Cadmium, or Lead chemisty batteries, Lithium batteries are have a lower weight per watt-second than the others.

2) Micro-Electro-Mechanical-Systems (MEMS) - which is a technique for building a mechanical system (like a balance beam) using the same processes that create integrated circuits. This opened up developing accelerometers, gyroscopes, and magnetometers where the sensor and the conditioning circuit were all in the same silicon die. That hugely lowered the cost of such things, and has evolved to the point of providing 9 degree of freedom systems that are on one, or two inexpensive chips.

3 - Cheap 32 bit Micros - Emergence of inexpensive 32 bit microprocessors with DSP like features. The ARM Cortex M series in particular. Even with a cheap inertial sensor you need to process it fast enough and with enough precision to act. DSPs can do this but they are complex, difficult to program, and development tools are expensive. 32 bit ARM processors are easily engaged by high school students using off the shelf free tools.

4 - High performance MOSFETs (low Rds(on) resistance), cheap hall effect sensors - these allowed people to build brushless motors with atonishing power to weight ratios. From CD-ROM spindle motors putting out 1/2 HP for electric planes to 15W motors the size of pager motors which are quite light weight.

Of course that all of this stuff is light weight gets the weight down to the point where you have enough power to lift it, and the integration gets the costs down to where you can build something on a small budget (a few hundred dollars, well within the budgets of active modelers)

isn't lithium more prone to failure on higher temps?

also, i'd love to know what the propellers are made of.

If it was me trying to film that close to a volcano, i'd simply use better lens :)

   > isn't lithium more prone to failure on higher temps?
Has not been my experience that this is the case, a low temperatures its storage capacity is reduced though.

Typical props are ABS plastic or Nylon. I suppose you could make them out of a higher performance plastic but I appreciate that they are cheap (I replaced a number of them learning to fly my quad copter)

The props are made of plastic (I have a similar unit as the person who made the video, I just don't live near a volcano). Coincidentally, I actually ordered a spare set today:

http://www.bhphotovideo.com/c/product/1013324-REG/dji_part3_...

Would you recommend DJI Phantom or are there other up and coming hobbyist drone manufacturers worth looking at?
It's still pretty new, but I have really enjoyed it. I have the Phantom 2 (not the Phantom 2 vision which comes with a camera) along with the gimbal and GoPro. The stability is astonishing. I have gotten a good hang of flying it, but haven't yet taken it on a trip so I just have videos of around town (in Florida, so there could be worse places to shoot in the winter).

It's still a little bit wild west in that you can tell the companies are still targeted at selling to RC hobbyists. The Phantom is good in this regard as it's ready to fly, but if you want first person view (FPV) that's still a chore (there are a ton of options and it's pretty much up to searching forums for people who have strong opinions on everything). I'd like to just toss some money at a company that provides a complete FPV solution along with up to date and easy installation instructions... Alas.

You could make one from parts. It's what I'm doing now.
It's a great way to go, if for no other reason than it will give you a stronger relationship with your aircraft and thus make you fly more carefully. :)

That said, while assembly and setup is mostly trivial (the hardest parts are basic soldering and dealing with some open source configuration GUIs for some flight controllers), figuring out what to buy and where to source it is pretty challenging, especially for a first build.

The cell phone industry has massively driven down the costs of the gyros and accelerometers needed to easily fly a quadcoptor. You also mentioned tiny cameras, another component that benefitted from the economies of scale of the cell phone manufacturers..

For example, you can buy a quadcoptor that fits int he palm of your hand for less than $40:

http://www.amazon.com/Estes-4606-Proto-Quadcopter-Black/dp/B...

By the way, that's the best toy quadcopter I've ever used (sample size of 4), and to say it fits in the palm of your hand is overstating its size. Three or four of them would easily fit in the palm of your hand, or the quadcopter and its tiny controller: http://cdn9.modelairplanenews.com/wp-content/uploads/2013/11...
I've got one of those too. I'm actually creeped out by how small it is, like a large dragon fly.
This is correct. In addition to the plummeting cost of sensors, communities such as DIY Drones and OpenPilot have made enormous contributions to the accessibility of this technology, by developing open-source autopilots that rival the capabilities of very expensive proprietary products.
the day isn't that far off where robots get set off to do work either hazardous, monotonous, or even just require constant precision, with regards to vehicle operation. As I posted before and I got the book idea from another here, Red Mars has a great amount of imaginative use of robotic vehicles and manufacturing.

Just like with this drone, besides eliminating some safety concerns we eventually will eliminate the other costly part, not having enough time to do all we want. With enough drones there won't be much we cannot do for time, just money will come up to the forefront.

Also the Robot series by Asimov, particularly Caves of Steel / The Naked Sun / The Robots of Dawn. http://en.wikipedia.org/wiki/Robot_series_%28Asimov%29
I'd say Lithium-ion batteries, cheap and power-efficient controllers, cheap sensors (orientation, acceleration, GPS), and the quad-copter layout.

With them you can build a device which is lightweight enough to fly for tens of minutes, and can easily orient itself. This lets you control it with high-level commands. You can say "fly there, face north, stand still", something you can't usually say to a traditional helicopter.

Exactly. Everything that the global smart-phone market pushed the prices down on.
Great answer(s)! (To you and other repliers -- thanks everyone.)

Makes sense. Not one single big thing (i.e. AWS for web companies), but a really great example of one industry benefitting and emerging because of cost improvements and advances in a different industry.

I, for one, continue to be excited about our drone overlords.

The multi-rotor drones are basically uncontrollable without computers; you need a reasonably powerful processor, which in turn means a fairly high-capacity battery. I think a lot of the innovation has come from mobile phone tech; there's now a lot more demand for small, light, high-performance processors, batteries and cameras, and so while much of the tech is not really new it's a lot more available to hobbyists than it was a few years ago.
The set of (say) 4 motors consumes much more power than the on-board computers. Brushless controllers these days handle 20 amps at around 5 volts for a total wattage of 400W(!) . The on-board computer power doesn't even count.

Of course, you're right that powerful computers are needed. But it's more about miniaturization for weight purposes.

Most fly with 8-bit Arduinos actually. Their power consumption is nil compared to the brushless motors.
> The multi-rotor drones are basically uncontrollable without computers; you need a reasonably powerful processor

True.

> which in turn means a fairly high-capacity battery.

Not true.

A microcontroller CPU is good enough. Take a look at the KK2.1 flight controller.

https://www.hobbyking.com/hobbyking/store/__49254__Hobbyking...

Its brain is an ATmega644, which only requires 240µA @ 1.8V, 1MHz, according to the spec sheet.

http://www.atmel.com/devices/atmega644.aspx

You do need energy-dense batteries, but all that energy goes into the rotors.

I agree with other posters that it has been the culmination of several things.

Batteries with high energy:weight.

MEMS sensors (thanks, nuclear weapon stewardship programs) http://mems.sandia.gov/

Cheap, open-ish, powerful SoC microprocessors, and the community of hackers (lots of help, and easy tools that lowered the learning curve).

Development of better "linear" circuits like advanced brushless motor controllers and chargers.

Growing robust hobbyist electronics tinkerer marketplace.

Maybe cheap 3D printing too, making prototyping easier/faster/better/more fun.

I think it's a combination of prices coming down for the quad-copters, cheap GoPro's providing HD video of the flights, and YouTube being available as a marketing platform to convince others to buy their own.