Hacker News new | comments | show | ask | jobs | submit login

Or they just don't want large hunks of metal with sharp spinning things possibly falling out of the sky onto people and things, and so would like to regulate the use of large flying things.

How about a mechanical fail-chute? The UAV could be equipped with a dead-man switch style parachute/inflatable life raft contraption that would trigger if the rotors stopped turning and if the device was flying high enough, it could float harmlessly to earth cocooned in a pocket of sturdy material buffered by some quantity of air. If you could figure out how to make it work, you could be in about as much danger as getting hit in the head by a beach ball...

Yeah, that sounds like a good idea. It's probably not going to be implemented across the board without some level of regulation, though. If they get to the point where they're passively safe, and require little skill to operate safely, the govt. might get to the point of fairly relaxed restrictions on them, like drivers licenses now.

This is not valid reasoning. Under this logic, we could feasibly be just as alarmed by balconies of any sort or platforms at any height greater than one story.

[Edit]: Which is not to say that there are hurtling objects on these things, but that this is talking about What Happens Least, and there is just as much risk of something falling from a balcony or second+ story area as there is a drone falling mid-flight.

There's a reason many sidewalks in NYC have sidewalk "sheds" covering them. Balconies usually aren't directly over places where people walk, for a reason. I'm guessing that this is due to laws.

To drop something really dangerous off of a balcony usually requires malice or gross negligence, and that generally goes straight down. The operator of a drone with crappy safety systems might simply have to leave it flying for too long without noticing the fuel gauge is nearing empty.

Once the FCC hashes out things like how drones should act automatically in safe ways such that untrained pilots can't easily kill multiple people with them by being slightly negligent or stupid, then I'm all for civilian drone flight.

I'm not sure I like the a priori nature assumed by the regulatory bodies. Why don't we legislate based on actual occurrences instead of just hypotheticals? I'm not sure why a law like this wouldn't work: "Any individual found to negligently operate a drone in a way that caused harm to persons or property will face jail time and fines".

Isn't that effectively the same as regulating the machine specifically, but without all of the overhead, corruption, and limitations on creativity, and wouldn't the law actually have the same effect if persons found to operate drones in an unsafe manner were actually prosecuted? Wouldn't the market only accept safe drones due to the possibility of jail time?

Just curious. Been kicking that line of thought around a lot as I've been in some creaky elevators recently.

As I see it, this is similar to food safety or mass-produced vehicle certification processes.

Unfortunately, modern markets are empirically fairly poor at optimizing for long-term outcomes with all externalities accounted for (example off the top of my head: the Deepwater Horizon oil spill).

When a major malfunction or accident does happen, no penalty or fine is really sufficient to bring about any meaningful sense of justice. (Think again the oil spill, or a tainted staple food poisoning tens of thousands of people, or to use a really cliché example, a malfunctioning UAV crashing into a group of kids playing at recess.) As the accidents are often due to incompetence or attempts at cost-cutting rather than outright malice, the deterrence factor of the penalty doesn't really apply – to those guilty, it just looks like they're taking the good kind of risk optimizing their systems until it's too late. The guilty corporation simply can go bankrupt and out of business, and finding and then charging personally responsible employees isn't always easy.

And yet Deepwater Horizon still happened. If the regulations that all rigs must use equipment bit x, y, and z, certified by industry conglomerate w, weren't in place, then maybe new, much safer technology would have been developed that would have made the spill not happen. If the market needs permission from bureaucrats, who may or may not be corrupt, or needs a new law passed, to adapt or develop better solutions to its problems, then it's no wonder we are stuck on obsolete tech that is 50+ years old and that the revenue in regulated markets is entirely dominated by a small oligopoly that holds everyone by the balls (see also phone companies).

Maybe, maybe not. What is the market motivation for developing and more importantly using technology which is safer on a 50-500 year scale? How high would you like the penalties for criminal negligence, manslaughter, etc, to be? Is it feasible for them to be higher than the potential savings on process optimization or cutting corners? Do you hold corporations or people, or both, responsible?

While government regulations might not be the most efficient, they do not cloud a picture of exactly how safe a plant is and will remain that would otherwise be perfectly clear. Short of each investor evaluating safety of each plant, oil rig, and factory themselves, the situation with government regulation removed would boil down to an analogy to the S&P credit rating business. We've seen how for-profit independent third-party rating business has worked out.

If you'd like an example from a market with little regulation, look no further than Bhopal.

I'm not going to argue that the current regulations we have in place are amazing. They're not and they can certainly be improved. I will, however, argue that they do a better job keeping us safer than the free market, as it has been implemented worldwide over the last century, would.

The FAA has always operated more like a white-list rather than a black-list. By default, nothing is allowed to fly in the national airspace. Then, regulations are added to allow a specific vehicle or a specific class of vehicles. At the moment, UAVs have not been white-listed, hence they are not legal.

Well that works fine until you're the one that's killed because of a failure. People after you are protected by the legislation passed, but you're SOL!

I don't think you understood -- I'm not suggesting we pass specific legislation detailing the cause of each accident after it happens. I suggest that we put in place generally applicable statutes for criminal negligence, manslaughter, etc. that would apply in case of malfunctioning industrial instruments. The law wouldn't change, but the market would automatically adapt and keep safe practices autonomously because the deterrent from the punishment was not worth the risk of cutting a few corners.

The difference is between being reactive and proactive.

Having a law in place is a start, but it has little effect against those who would cut corners. I believe the saying is that "laws only keep the lawful honest." There will be a lot of manufacturers building TacoCopters who will cut corners because it's cheaper, and others who simply don't care because the don't expect to get caught.

However, if the FAA mandates (e.g.,) complete engine teardown and rebuild every 200 operating hours along with an annual FAA inspection, they can actually verify that those procedures are being followed. The law will still be in effect, but verifying that it is being followed happens well before someone is killed and will take the vehicles most likely to kill someone out of the air.

That's what they're trying to accomplish.

>This is not valid reasoning. Under this logic, we could feasibly be just as alarmed by balconies of any sort or platforms at any height greater than one story

This is why we have building codes and regulations. Since the FAA did not have these regulations in place, it was disallowed (commercially.) Congress recently said "make the regulations" because it seems like there is enough demand to justify the increased burden on the taxpayer.

Which we are by way of building codes and regulations...

That's exactly what I'm getting at, here. The accusations of this being because there is an ominous police state is ludicrous when we could just put expectations of responsibility in place.

Apart from other good points mentioned in comments, there's also this thing that in free-fall scenario (h = how high you object is above the ground, v = velocity, Ek = kinetic energy),

v(h) ~ sqrt(h)


Ek(h) ~ v^2 (h) ~ h

Which means, the higher something starts, the more it will hurt when it hits your head.

They're perfectly happy for the police to use these drones so that's not the problem here.

Who is perfectly happy?

Police in Mesa County, Colorado received the first authorization ever for police to fly UAS. That was last year. Since then, a county in Florida received authorization. Police in Texas toyed with the idea, but decided it wasn't in the budget. So i know of precisely two police precincts with legal authorization. It's possible there's a couple others I haven't heard about, but the total number is still very small. Even the Utah Highway Patrol got slapped down by the FAA for flying without authorization.

Have the DENIED any requests from law enforcement? No. Perfectly happy.

Can you substantiate your claims? I can.

M. Wilson at the FAA has denied applications for certificates of authorization (COAs). Without his signature, the popo doesn't get to fly.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact