Packs of robots will hunt down uncooperative humans

Short Sharp Science (NewScientist blog) Oct. 22, 2008

The latest request from the Pentagon jars the senses. At least, it did mine. They are looking for contractors to provide a "Multi-Robot Pursuit System" that will let packs of robots "search for and detect a non-cooperative human".

One thing that really bugs defence chiefs is having their troops diverted from other duties to control robots. So having a pack of them controlled by one person makes logistical sense. But I'm concerned about where this technology will end up.

Given that iRobot last year struck a deal with Taser International to mount stun weapons on its military robots, how long before we see packs of droids hunting down pesky demonstrators with paralysing weapons? Or could the packs even be lethally armed? I asked two experts on automated weapons what they thought. Both were concerned that packs of robots would be entrusted with tasks - and weapons - they were not up to handling without making wrong decisions.

Steve Wright of Leeds Metropolitan University is an expert on police and military technologies, and last year correctly predicted this pack-hunting mode of operation would happen. "The giveaway here is the phrase 'a non-cooperative human subject'," he told me:

"What we have here are the beginnings of something designed to enable robots to hunt down humans like a pack of dogs. Once the software is perfected we can reasonably anticipate that they will become autonomous and become armed.
We can also expect such systems to be equipped with human detection and tracking devices including sensors which detect human breath and the radio waves associated with a human heart beat. These are technologies already developed."

Another commentator often in the news for his views on military robot autonomy is Noel Sharkey, an AI and robotics engineer at the University of Sheffield. He says he can understand why the military wants such technology, but also worries it will be used irresponsibly.

"This is a clear step towards one of the main goals of the US Army's Future Combat Systems project, which aims to make a single soldier the nexus for a large scale robot attack. Independently, ground and aerial robots have been tested together and once the bits are joined, there will be a robot force under command of a single soldier with potentially dire consequences for innocents around the corner."

What do you make of this? Are we letting our militaries run technologically amok with our tax dollars? Or can robot soldiers be programmed to be even more ethical than human ones, as some researchers claim?

Paul Marks, technology correspondent