Ask HN: What are the major problems in drone technology?
Is it... Height? Speed? Range? Battery life? Lack of autonomy? Size? Reliability? Other?
In particular, how are these limitations hurting potential applications?
In particular, how are these limitations hurting potential applications?
[0] For flying types:
- Regulation. Which is understandable, due to the safety risks of autonomous flying machines.
- Energy storage/consumption. Electric drones are the most common now, but there could be a move towards Petrol based ones once technology catches on (lighter/more efficient engines).
For land type:
- Mobility. Right now there is a lot of focus on leg-based and weg-based[0] designs.
- Vision. Object recognition is slowly coming along, but the POV of a land based robot makes it a bit more difficult.
- Costs.
For water type:
- Costs.
- Regulation.
[0]From my experience. [1]Wheel-leg hybrid design.
I can't find the original article I saw this on but it seems like one of those things where even though the big market is probably smaller hybrid cars, might have application in commercial drones since it can potantially save a lot of weight vs more conventional engines.
[0]http://www.osengines.com/engines-airplane/osmg1401/ [1]http://rcvehicles.about.com/od/nitro/f/nitrofuel.htm
From a consumer standpoint, the pricing has come down substantially. You can build a fully autonomous quadcopter for ~$500 these days (minus the transmitter).
The problems I see are regulation, and battery performance.
Regulations are really up in the air right now. There have been two rulings by judges that the FAA's current policies do not cover restriction of commercial usage for drones. The FAA is actively fighting this. I feel that over the next few years, we'll see some solid rules set down, and more than likely they'll be extremely strict to start. Statistically speaking, the FAA has an amazing track record, and letting anyone into the skies has a real chance of marring that record substantially.
Cell power density is getting better and better, but it still takes a lot of work to get over an hour flight time on a multirotor. Thus they tend to be used for on-spot and precision flying shots for video. Fixed wing drones on the other hand have amazing range (I've seen 70km+ flights), but are less capable in the precision flying that a multirotor provides.
On this topic, I am a huge fan of what Chris Anderson is doing over at 3DRobotics. Their open-source APM is the flight controller I choose for most of my projects, and it is extremely configurable.
Hope this helps :)
On the technical side we faced a few chanllanges. The first was reliability. Our planes were small <10 lbs, but would enevitly go down and be very difficult to recover. After the first couple crashes we designed a protective shell around the expensive bits. Causes of the crashes we always unique, from a loose connection, to interference.
The next biggest problem was battery life, ours was an electric system and getting more flight time in the tiny platform was difficult. This of course limited the range which limited it usefullness.
Otherwise, we solved the autonomy with off-the-shelve components (albeit military grade). Otherwise most other issues were resolved with software (flight planning, data recovery, etc).
Our application was aerial imaging, I imagine other applications would have another set of problems.
If you consider a city like Los Angeles, with 3792621 inhabitants, and a surface of 1215e6 m², assuming each person takes 1 m² (ie. the person-drone cross section), that gives for fall points equidistributed over the surface a probability of hitting a person of 3792621/1215e6 = 0.0031; only 0.31% per fall! To hit a person with 1/2 probability, you'd need a rainfall of 160 drones!
Will your drone fly over crowded areas? Most people are inside houses most of the time anyways, and there's no reason to fly over the sideways.
Now, you must also consider other areas to avoid landing (or falling) on. Cars, (streets), random machinery (eg. air conditioners), power lines, etc. But all in all, that still leaves a lot of places where to crash with no consequence, if you need to crash.
IMO, there's a lot of unreasonned fears about the subject (as there is about anything else; and people wonder in other threads, why the belief that humans use only 10% of their brains persists...).
And kudos on your calculations, I had never seen this done before. This is the type of math that "news" stories need to have.