US Department of Homeland Security looking for (more than) a few good drones • Hacker News Board

Re: US Department of Homeland Security looking for (more than) a few good drones

#31

Post by telecuda » Sat, Oct 06, 2012, 9:17 PM UTC

I was at a law enforcement show this week and drones were the big new thing. Most were either quadcopters with GoPros while others looked like mini helicopters with PTZ cameras strapped to them.

Better funded departments are starting to test them out, but the demand from local police still isn't quite there. They require obtaining a special FAA license and lots of training, which makes the sales cycles on drones long and unattractive to vendors selling one at a time to local PDs.

Still though, their pitch is: "A helicopter flight cost is $450/hr. A drone is $14." Saving PDs the cost of flying a drone to/over an accident scene is more the draw than snooping into buildings.

Re: US Department of Homeland Security looking for (more than) a few good drones

#32

Post by mmagin » Sun, Oct 07, 2012, 3:30 AM UTC

revelation wrote: post #11

sturadnidge wrote: post #5

I was watching a documentary the other day about border smuggling - surely it's only a matter of time before drones are used for that, in which case what hope has domestic law enforcement got if they don't go down the same path.

Smugglers have used submarines. Law enforcement has already "no hope", and that is entirely by design. If you want LE and criminals to be equal then you end up in a country run by criminals.

No, there is more than the technical advantage held by police which factors into this. There is also the extent to which the mass of people tend to believe whether it is better to do things in a law-abiding manner or whether legal authority is completely disrespectable and not worth obeying.

Re: US Department of Homeland Security looking for (more than) a few good drones

#33

Post by ktizo » Sun, Oct 07, 2012, 10:06 PM UTC

starpilot wrote: post #28

ktizo wrote: post #10

The scariest thing about drones is that if both sides have drones, the winning strategy would seem to be in having millions of cheap drones directed by the sort of algorithms currently used in high frequency trading. Putting humans in the decision making loop for all of those just slows them down and vastly increases the expense and manpower requirements.

There's very, very little work going on with AI in drones. We're not there yet. We're still working on getting the sensors right and dealing with datalinks. The vast majority are remotely piloted, some are remotely directed (point and click to location, the drone maneuvers to get there), all are controlled by humans. This "robot drone aircraft freakout" sounds as silly to me as an aerospace engineer as worries about…

I am not pointing out current capabilities, I am just hypothesizing on what is likely to develop in drone war when more than one side has drones.