New mobile-payment technologies like Apple Pay and Samsung Pay require a fingerprint to authenticate the user, but their biometric scanners may not be accurate enough to bet your bank account on. Mobile fingerprint readers such as those in the iPhone and the Galaxy S series are vulnerable to fake fingers and, when they work, often make you swipe multiple times before recognizing your input.
The problem, according to Vkansee CEO Jason Chaikin, is that the haptic sensors in the Apple and Samsung devices only capture 500 dots per inch, not enough resolution to distinguish a mold from a real human finger. At last week's Mobile World Congress trade show in Barcelona, Spain, Chaikin and his team showed off UTFIS, a mobile fingerprint scanner, which uses an optical sensor to get a 2,000 dpi image, complete with sweat pores and other fine details that today's fingerprint readers miss.
As I watched, a company technician took a rubber fingerprint mold and pressed it down against the sensor on an iPhone, which unlocked immediately. Clearly, if a thief can steal both your phone and your fingerprint, it would be easy to gain access to the device and start making purchases in your name.
I also took a closer look at the swipe sensor on Samsung's Galaxy S5, which is an annoying device because it forces you to move your finger over the bottom of the phone at just the right speed. If you move your finger too fast, or at the wrong angle, the reader won't see enough of your finger to identify it, and you'll have to swipe again and again until you get it right.
Vkansee Software Engineer Ernst Mucke explained that high-security fingerprint readers such as those employed by the U.S. government always use optical scanners. However, phone makers like Samsung and Apple use less accurate haptic or swipe sensors because those sensors are thinner and less expensive.
UTFIS promises the best of both worlds. It is just 1.5mm thick, but scans at four times the resolution of today's mobile fingerprint readers. It achieves its svelte form factor by using a pinhole design rather than a full-lens system.
When I held my finger against a sample UTFIS scanner which was attached to a PC, I saw a black-and-white video feed showing a very detailed view of my skin, including all the ridges and pores. Mucke said that the detection software is smart enough to recognize your finger in different temperatures, and even when your finger is wet, although both temperature and moisture would change the look of your ridges and pores slightly.
Vkansee plans to demonstrate its UTFIS fingerprint scanner to manufacturers this spring. With luck, the next generation of smartphones will do a better job of recognizing your fingers than today's does.