Sunday, March 06, 2016

Enjoy it while you can

Since we live in an umpteenth generation warfare world, what do autonomous drone aircraft/submarines mean to us in everyday life?

I know of three classes of drone aircraft: slow consumer loitering-capable, slightly faster consumer non-loitering, and high performance military gear. There’s probably plenty in between I don’t know about, and if there isn’t now there will be if somebody sees a need.

I assume jamming will work fine against completely remote-controlled drones, so fixed sites (isolated military bases, etc) can protect themselves pretty well. Not so random civilians.

Obviously drones are handy for assassinations. Top figures (e.g. heads of state) probably get pretty good drone spotting and interdiction coverage, and if not they will soon. Threats to the rest of us could be serious, though. Consider a loitering drone, completely remote controlled, that waits for the police chief to take out the garbage. Or if he’s willing to assume that kind of risk, waiting for one of his kids to appear. Yours truly thinks it reasonable to automatically shoot down drones inside city limits, whether Amazon likes it or not. And yes, city defenses would have to get waivers from the Migratory Birds Treaty Act—you couldn’t help nailing a few.

Most of us aren’t, outside of a civil war, likely to be assassination targets—and during a civil war partisans would probably run out of drones pretty quick and rely on the two-legged varieties.

Consider a military airbase inside the USA or friendly country. When they are far from cities, the approaches are over easily controllable area that can be laced with whatever detection and interdiction hardware they please. Non-existent neighbors won’t care about spectrum jamming or the occasional rocket landing on the roof.

We already have technology for trying to spot and scare off birds. This can be beefed up and given some teeth. Slow drones are being attacked with net guns; I assume automatic shotguns would work too. Higher altitude drones are harder to get at. Using one’s own patrol of interceptor drones might be the solution. Note that this works against autonomous drones as well as remote-controlled. Even so, there would be some non-zero breakthrough rate, and one could always overwhelm the defense somewhere.

Civilian airports, or military/civilian, or military that suburbia has surrounded, are a different problem. They can’t arrange for defense in depth. What could happen to them?

One simple attack is to have drones with explosive payload home in on the biggest radio signals. This could knock out jamming, but could also take out the local radar. Semi-autonomous drones could also damage the radar system, and in those places that still use a traditional control tower, make hash of that also. That makes the facility harder to use.

Hitting an aircraft doesn’t sound terribly easy, but loitering drones could stack up (no, not right over each other) in the descent path and let an aircraft hit them. The closer in to the landing field, the shorter the window the pilot has to hit, and the more likely your two dozen drones are to intersect the aircraft.

You could knock drones down if a fast jet buzzed the flight path, but that’s kind of risky for the buzz-er and leaves the air pretty turbulent for the next plane trying to land—another big risk. Maybe a few helicopters with nets could try to wrangle the drones? The next generation of drone software could include “big slow thing evasion,” which wouldn’t get them out of the way of jets but would with helicopters.

In any event, you can see that landing passenger jets every 90 seconds starts to look hard.

So, how do you deal with this sort of problem? The safe solution is to abandon the fixed investment in the existing airports and move them to the boonies, with fast transport to and from. Airlines would scream bloody murder. (And, as best as I can tell, the military bureaucracies aren’t always very good at hard choices either.) The usual approach (clamping down on sales of whatever technology) isn’t likely to be of much use, but it will be tried.

So, air travel becomes slower and more expensive, and more risky.

OK, how about sea travel? Autonomous loitering drone submarines could be coordinated to attack enemy ships, provided communications weren’t jammed. Suppose communications are proactively jammed. Drones would have to be able to ID their targets, and would tend to act alone, though possibly launching sub-drones (aka torpedoes?) to attack from multiple directions.(*)

Suppose ID isn’t easy, or this is a chunk of the ocean where you don’t care who gets hit. Daesh et al don’t much care who they hit. Then what you create is like a mobile mine—search out anything big and attack it.

Last time I checked, that kind of risk made transport costs soar. Luckily, loitering drone submarines are, and are likely to remain, expensive. Mobile mines could be quite a bit cheaper, but moving through the water takes a lot of energy, and moving on the surface exposes it to detection.


How about trains? If you armored them a bit? A drone with explosive payload can always take out a track, and if it is timed right so the train can't stop you could derail it. (I assume the tracks are instrumented, or will be.) Countermeasure: armor the train and slow it down, and add some drone knock-down flanges on the side (oopsies going through tunnels...).

Yes, of course I am assuming that some of the enemy will have infiltrated, and will not distinguish military from civilian targets. Why would they do things the hard way?

(*)I figure that sea battles will be pretty complicated. Where are the enemy drones, where are yours, how do you coordinate search and destroy, how do you keep your own screen of defenders—all in seawater and bad weather. With both aircraft drones and submarine drones, all partly autonomous. Sort of like fleet battles, but with more moving pieces.


Assistant Village Idiot said...

This is not at all comforting.

How traceable are drones that shoot down a police chief?

james said...

I assume that if they are completely controlled, the controller needs to have near line of sight to the drone. Buildings can shadow higher frequencies. And there's lots of noise. Worth looking into.
Buildings can also produce reflections that could make triangulation a bit trickier, but if you instrumented every other block or so you could probably find the source. Some cities use similar instrumentation to listen for gunshots.

Semi-autonomous drones are a harder problem.