Comments on: UAVs (Follow Up)

By: demophilus

demophilus — Wed, 19 Sep 2007 18:49:17 +0000

Agreed. Didn't mean to suggest that a carrier UCAS is invalid. Just don't think it'll go off without a hitch. That's not how shit happens. Don't think we've seen the end of man in the loop either, and I don't buy "no human input or interaction to complete their missions". In war things don't go according to plan. If Plan A is a machine, Plan B is having a human in the loop to reboot or troubleshoot it. That's simple redundancy. Didn't mean to suggest anything more.

By: Robophilus

Robophilus — Wed, 19 Sep 2007 12:15:28 +0000

Demophilus, human pilots are not “foolproof” either.
A lot of us don’t trust human pilots. They crash, sometimes for no reason. In the past 100 years, humans have broken a lot of airplanes trying to get them on and off aircraft carriers. Yet that doesn’t invalidate the “human pilot” concept — nor should (inevitable) accidents invalidate the robot pilot concept.

By: Trent Telenko

Trent Telenko — Tue, 18 Sep 2007 22:16:09 +0000

I think this passage from a August 20, 2007 Financial Times article titled “US Military In Dogfight Over Drones” covers the UAV issue from the US Army point of view.
>

By: demophilus

demophilus — Tue, 18 Sep 2007 21:38:43 +0000

Rip:
Maybe USAF’s figuring that if the USN/carrier system works, they can dump the arresting gear, lighten the landing gear, and get a more capable dual mode system. Or maybe that’s just a taxpayer’s fantasy.
J.A.:
No offense, but it wasn’t so long ago that the F-22’s computer systems crashed on their deployment to Okinawa. The meat puppets had to fly them back to Hawaii.
A lot of us don’t trust our computers. They crash, sometimes for no reason. Especially, if we’re running Microsoft apps.
Redundant, highly tested systems are a beautiful thing, but sometimes they’re only as foolproof as their assumptions. Catastrophe theory and uncertainty are the enemies of anything highly engineered, and they’re hard (if not impossible) to beat. You can test something until Doomsday, and all you’ll find is what you tested for. Some glitches only happen later. When they do, they can be spectacular.

By: Rip

Rip — Tue, 18 Sep 2007 16:27:22 +0000

Looking at the big picture, it appears the USAF made a decision to withdraw from the UCAV track for publicly vague reasons. Some believe it was a victory for the “fighter mafia” insuring much less likelihood of competition for the F-22 and eventually, the F-35. The hard fighter jocks look at the 2001 NDAA and say, “F’em — they’re just a bunch of politicians…“
They (USAF) are still advancing the ISR UAV (Preaditor/Reaper, Global Hawk, and perhaps some black programs); however, the increasing success of the armed MQ-1/MQ-9 is an embarrassing reality they would like to go away. (It will not!)
Meanwhile, the recently floated USN studies make a very compelling case for the CAG UCAV to insure the carrier’s relevance in at least the first half of this century.
The next key strategic event will be if the USAF proposes as part of the Global Strike System (i.e., the B-3), to include an unmanned MQB-3 variant. If they don’t elect this option, and have nothing under the black kimono, then they are ceding the technology and lead to the USN for years to come.
The “man overseeing the loop” is the future in the DoD, it’s sad to see the USAF, who toots the image of being in tune with the future, fighting a battle based on tactical retrograde.

By: cynicism

cynicism — Tue, 18 Sep 2007 12:56:37 +0000

There is absolutely no chance of a bug or an AI “glitch”. Zero. And anyway, if it hits the wrong target, this is Iraq we’re talking about.. What’s the big deal?

By: JUCAS Architect

JUCAS Architect — Tue, 18 Sep 2007 11:45:23 +0000

This article is FUD. First and foremost, “hiccups in the data stream” don’t happen with an autonomous vehicle. The X-45 and X-47 vehicles aren’t dumb fly-by-wire systems like Predator or GlobalHawk. They are completely autonomous vehicles that are in control of their entire flight regime from take-off to landing and require no human input or interaction to complete their missions. Second, the ability to auto-land on a carrier has existed for decades. That it isn’t used or widely known is a testament to the machismo of Naval aviators and not an indictment of autonomous landing technology. Anyone espousing fear of this technology is expressing a profound ignorance of the technical realities.