Two articles about advances in consumer-focused technologies have caught my eye recently. While neither was written with the theme of physical security, I nevertheless saw their topics as potential game changers in our market: Wearable cameras and software that tracks facial expressions to predict future behavior.
The Wall Street Journal profiled two photographic cameras — the Narrative Clip (getnarrative.com) and the Autographer (autographer.com) — that capture a continuous stream of images while being worn around the neck.
The $279 Narrative Clip is a 5 MP camera which takes 2 geo-tagged photos per minute; has GPS, accelerometer and magnetometer sensors, USB connectivity, 8 GB of memory and two days of battery life. Its current capability is two images per minute. The $399 Autographer is also a 5 MP camera with fisheye lens, accelerometer, compass and thermometer sensors, USB and Bluetooth connectivity, mobile app and 10 hours of battery life. The Autographer can be set to shoot nine consecutive images, although I couldn’t find a specification for the time span of these images.
The WSJ article focused on issues of privacy and proper etiquette, such as actually informing people they are being photographed. Concerns like this are probably more applicable to pendant-type devices than the much-publicized Google Glass, simply because the Google device is a lot more obvious — if you are looking someone in the eye, you are looking at the device.
However, in security and adjacent applications, I think pendant or other semi-obvious devices could have some real potential, particularly in customer-facing situations, such as banking and retail. Here, a closer view or better camera angle could provide a better image than surveillance video, without the action of holding a camera up to the eye. Security officers at checkpoints and security guards might also be provisioned with these to record persons of interest, situations or transactions. With increased storage and battery life, video versions of these devices are already popping up — mostly targeted at the law enforcement community, from vendors such as Digital Ally (digitalallyinc.com), Vievu (vievu.com) and others, opening up even more possibilities.
Miniaturization, low cost, compact storage, and battery life are all unmistakable trends which will affect the way we deploy and use image technology, including thermal. Further, we have consistently seen the consumer market lead the security market by bringing technology into the affordable range through volume scaling of technology. If all the hype around wearable technology at January’s Consumer Electronics Show is to be believed, surely the security market will find a way to use it.
Security Meets Social Science and Consumerism
The other piece I found interesting was published in the January issue of MIT’s Technology Review. It describes an algorithm developed by a Massachusetts startup company named Affectiva (affdex.com). The algorithm tracks facial expressions — raised eyebrows, frowns, smiles, etc. — and processes the data to predict future behavior. The product, called Affdex, is being developed as a “neuromarketing” tool, the company’s term for a marketing and sales prediction tool.
Once again, it seems that security applications may be ripe for this type of technology. Could this potentially detect and process expressions of stress, fear or mal-intent? Could it predict shoplifting behavior, pre-robbery jitters, credit card fraud or an impending terror event? Would this be of use to TSA at its airport checkpoints, to stadium and arena security personnel, or at high-security facilities?
Behavior and behavioral analytics remains one of the most significant areas for advancement in predictive or proactive security. This particular technology works with standard webcams and an SDK is available for custom applications, so I suspect someone will investigate it for suitability in security event prediction. Combine developments such as this in the behavioral arena with continuing advances in facial recognition and big data, the result could be very powerful predictive and analysis tools.