As discussed here, scientists at Shinshu University in Nagano, Japan were able to correctly identify people by their footsteps 99.8% of the time.
Previous studies only sampled ten people at a time. This study had 104 people walk across a half-meter-long board containing thousands of pressure sensors, recording ten steps per person and how each person’s foot applied pressure differently with each step. The measurements (having been processed by a unique algorithm) proved to be rather unique. Of over 1,000 steps recorded, the computer only misidentified three steps.
The scientists envision this technique being used in airports to identify passengers. But if U.S. passengers might soon not have to take their shoes off at airport security, I do not think they will be excited about the prospect of walking barefoot across a sensor board (especially after thousands of others have done so that day).
Of course, various means of “gait analysis” have been proposed before. But these systems are designed to identify individuals by the way their entire body moves while walking, not just how pressure is applied by a foot during a stride. In Little Brother, Cory Doctorow wrote about a student who filled his boots with pebbles to intentionally affect his gait and confuse the sensors.
But this study is unique in that it was able to sort out a much larger sample after only ten steps per person. As sensors become more widespread, and as computing power becomes faster, cheaper, and smaller, such sorting and identification is likely to become more common in the future. What might be the consequences of such developments?
In order to use this system, I assume passengers will have to be measured and recorded prior to passing through the system. What if this new metric was necessary to get a “flyer’s license”? What if it was recorded when you applied for or renewed your driver’s license? Who would store these records? How would they be shared across all airport security systems?
Where humans can be identified, they can be recorded, tracked, and analyzed. What if the system improves so you don’t need to be barefoot? What if these systems are installed at grocery stores, malls, or the workplace? What might such a system learn about you?
This is but one of many ways humans can be identified, and new ones are being developed as we speak. What if, in a post-singularity society, computers are able to develop new ways of identifying and tracking us? What if humanity is not aware of these methods? How would we counter them? Would we want to?
If nothing else, this study should give you something to think about while you stand inside the airport backscatter scanner. What do you think of the study? Where are we headed with this software?