Running biometric footware

There is a new biometric concept afoot from a Vancouver startup, turning a homespun invention for comfortable shoes into an entirely new source of biometric data.

At the moment, the pre-funding company is little more than a weirdly graphical but vague website (http://www.plantiga.com/), some patent applications, and the gushing energy of founder Quin Sandler, who nonetheless ties himself in knots trying to chat up his technology without actually saying anything specific about it.

What is this new tech? Sandler asserts that it is a "changing concept adaptable architecture" for constructing shoes, but then backtracks (on legal counsel) to describe only an "enclosure for interlocking structures." No, let's try "components which create an adaptable interface between floor and person." Yeah, that's it: a new way of making shoes. Oh, and extracting data from them.

Only they haven't made any shoes yet, nor extracted any data. As Sandler says, "We're above concept, but below prototype... we believe it's cool, but we still need to make it," presumably after they get the first one or two million dollars.

That money will create the shoe itself, a comfy and cushy distributed shock-absorber which for secret reasons will be more comfortable than what's on offer now. But that's a lot to spend on a new but distinctly low-tech shoe, without actuators or adaptive algorithms. How does that traditional footwear become high-tech "footware?" (And does it need to reboot?)

The brand-new piece is what Sandler calls "gait biometrics," a fancy term for identifying an individual based on signals transmitted from his feet. The idea is that those mysterious "components" of the shoe somehow transmit real-time pressure signals to the outside world, and those form the basis of a new dynamic biometric profile. Rather like voice-printing, or dynamic signature capture.

Each person's feet offer a different signal. Supposedly, anyway, since Sandler doesn't have any research to back up the claim that any person's gait is constant enough to serve as a profile, and different enough from others' to uniquely identify him. He hasn't actually logged the data yet, nor transmitted it wirelessly, and certainly hasn't worked out the algorithms to digest it.

He freely admits the biometric development is "very immature," with "no mathematical development done." But he's still excited that his approach uses a continuous stream of real-time data, as opposed to the old-fashioned, fuddy-duddy "single point of control" systems like fingerprints and iris-scans.

Sandler apparently dreams of elaborate but unspecified algorithms which somehow extract possibly-existing patterns from hypothetical data; sure beats real engineering. However vague and improbable, this always-on biometric concept makes for some catchy slogans: "A unique form of physical access control", and "The environment knows WHO YOU ARE and WHAT YOU ARE DOING."

The hype apparently works: Sandler just got a call from the Pentagon's skunk-works research agency DARPA. One small step for the jackboots. ®

Bill Softky has written a neat utility for Excel power users called FlowSheet: it turns cryptic formulae like "SUM(A4:A7)/D5" into pretty, intuitive diagrams. It's free, for now. Check it out.