Skip to content

The Underground, Viewed from Above

September 22, 2011

Picture a future in which the earth’s crust has been totally interpenetrated by teeming swarms of autonomous, tunneling robots.

The system is an innovative all-terrain mobility platform capable of accessing tunnel systems through a small (8 inch) borehole and traversing adverse tunnel terrain including vertical obstacles up to 2ft in height and chasms up to 2ft in length.  The system’s function is to provide a platform capable of carrying a small sensor package while navigating and overcoming terrain obstacles inside the tunnel.  Counter tunnel technologies are needed to support intelligence gathering and safety of troops and personnel in unmapped and unknown tunnel environments.  The system is the initial step in achieving a fully autonomous counter tunnel system. – Innovative Solutions for Counter-Tunnel Robotic Surveillance

But, as BLDGBLOG notes, low ground is the new high ground, because this thing is being run under the auspices of the USAF:

there is perhaps also an indication here that a conceptual revolution is underway within the Air Force, wherein the earth itself—geological space—is seen as merely a thicker version of the sky. That is, the ground is now seen by Air Force strategists as an abstract, three-dimensional space through which machines can operate, like planes in the sky, navigating past “terrain obstacles” like so much turbulence. In a sense, the inside of the earth becomes ontologically—and, certainly, technically—identical to the atmosphere: it is an undifferentiated space that can be traversed in all directions by the appropriate machines.

Flying and tunneling thus become elided, revealed as one and the same activity; and the Air Force is understandably now in the business of the underground. – Tunnel / Countertunnel

This is a special case of a more general trend. From the point of view of the systems designer, all spaces are collapsed onto a single homogeneous substrate, with relevant details and feature vectors selected algorithmically and then kicked upstairs to be used as input by a human or machine decision-maker. The conceptual space of a machine-learning algorithm (if you’ll pardon the premature anthropomorphization) is unbounded and uniform. This is another sense, now a topological one, in which everything is connected.


From → Uncategorized

Comments are closed.