Subterranean mole-bot under development

No plan to seek out underground civilisations, though

British boffins are developing a mechanical, mole-like robotic system which could dig its way through debris or rubble. The technology could be used to build burrowing rescue machines able to retrieve people trapped by collapsing buildings or tunnels.

New Scientist reports that the men behind the new kit are Robin Scott and Robert Richardson of the Artificial Intelligence Group at Manchester's School of Computer Science. The engineering brainboxes were apparently inspired by the digging action of talpa europea - the common mole - rather than the fiction of Edgar Rice Burroughs et al.

The two boffins have been working on mole-machines for some time. In this paper dating from 2005 they point out that: "Digging involves subterranean tunnelling or burrowing activity... This area of [robotics] research has been extremely neglected."

We at the Reg - well, some of us anyway - strongly agree. We don't have our flying car, nor our robot/brainchipped-monkey butlers. On top of that, we've yet to hear of the long-promised underground mechanical mole cruiser, able to roam about beneath the Earth's surface without leaving a tunnel, spoil heap, and power connection behind it. Even a relatively unimpressive subterranean droid would be nice.

Well, actually we have seen that last one (pdf). But today's mole-derived kit seems more intuitively sympathetic than existing wormoid proposals.

Unfortunately, the Mancunian mole-machine isn't a complete system. So far, it's no more than a digging mechanism, and the university's test videos are rather lacking in wow factor.

We're particularly unimpressed with the clingfilm taped over the machinery for the digging-through-smaller-lumps bit. Still, it's a wriggle in the right direction. Make it a lot bigger, add some kind of miracle sensor/navigation gear and hugely puissant air-independent power source (nuclear? Steampunk HTP?) and we're on. Intercontinental subterranean mole dreadnoughts are us.

Sadly, solving those other problems of power and sensing will be hard to do. Merely copying animal locomotion techniques is a fairly trivial activity by comparison.

Still, Scott and Richardson recently told New Scientist that they're moving forward to proper mobile system tests, and that they might have a working rescue machine good to go in two years.

It's always possible that there's some real engineering going on here, not merely more robotics-faculty copying of ideas from biology textbooks. ®

Sponsored: 10 ways wire data helps conquer IT complexity