Robotics is finding more and more use on the battlefield, so it's logical for DARPA to push the state of the art of one particular aspect of the technology. I'm sure there are other programs to deal with the munitions aspects.
Back in the early eighties I worked at major defense contractor. This was during WWIII, aka the cold war. One of the projects (that I did not work on) was an autonomous anti-tank weapon. The fear was that the Ruskies would roll across western Europe led by vast columns of tanks. The US wanted a non-nuclear response so as to avoid the use of tactical nuclear weapons.
The idea was to fly over likely routes for the tank columns, and drop automated weapons in vast quantities using parachutes. Upon landing, the weapons would deploy an array of microphones and a ground-coupled accelerometer. A microprocessor on board examined the inputs to detect vehicles, determine their type, estimate and project their path, determine the optimal intercept point for a heat seeking munition, and launch that munition.
Another SW routine detected the footfalls of an approching person, and would self-destruct -- taking out that person. It all seemed very abstract to me until I heard about that particular feature. Taking out the diesel engines of tanks and trucks seemed somewhat benign (if one blindly ignores the possibility of incinerating a tank crew), but writing software to purposely shred some poor Russian farm kid seems rather cold.
Having a measures-countermeasures mindset, I thought the Russians should build little robots that would fling themselves on top of the automated anti tank weapons. (The Iranians chose the low-tech approach of using swarms of human minefield sweepers during the Iran Iraq war.)
The idea of battle bots fighting battle bots has a certain appeal -- bloodless conflict. But this is an illusion, since battle bots are ultimately used as a better way to kill.
|
Back in the early eighties I worked at major defense contractor. This was during WWIII, aka the cold war. One of the projects (that I did not work on) was an autonomous anti-tank weapon. The fear was that the Ruskies would roll across western Europe led by vast columns of tanks. The US wanted a non-nuclear response so as to avoid the use of tactical nuclear weapons.
The idea was to fly over likely routes for the tank columns, and drop automated weapons in vast quantities using parachutes. Upon landing, the weapons would deploy an array of microphones and a ground-coupled accelerometer. A microprocessor on board examined the inputs to detect vehicles, determine their type, estimate and project their path, determine the optimal intercept point for a heat seeking munition, and launch that munition.
Another SW routine detected the footfalls of an approching person, and would self-destruct -- taking out that person. It all seemed very abstract to me until I heard about that particular feature. Taking out the diesel engines of tanks and trucks seemed somewhat benign (if one blindly ignores the possibility of incinerating a tank crew), but writing software to purposely shred some poor Russian farm kid seems rather cold.
Having a measures-countermeasures mindset, I thought the Russians should build little robots that would fling themselves on top of the automated anti tank weapons. (The Iranians chose the low-tech approach of using swarms of human minefield sweepers during the Iran Iraq war.)
The idea of battle bots fighting battle bots has a certain appeal -- bloodless conflict. But this is an illusion, since battle bots are ultimately used as a better way to kill.
- mark 2-27-2004 2:08 am