Researchers at the University of California, Davis and Berkeley have managed to miniaturize medical ultrasound technology to create a fingerprint sensor that scans your finger in 3D. This low-power technology, which could improve on the robustness of current-generation capacitive scanners, could soon find its way to our smartphones and tablets.
When Apple announced it was going to include a fingerprint scanner in 2013's iPhone 5s, questions immediately arose as to just how safe and accurate the scanning would be. As it turned out, the scanning was accurate enough (unless you've just gone swimming), and the feature, which was generally well received, soon found its way to many other smartphones.
But even in the best of cases, the capacitive sensors used in the current generation of portable devices are still subject to serious security leaks. These scanners only image your fingers in two dimensions, and so they are easily fooled by placing a printed image of a fingerprint on top of the sensor.
Professor David Horsley and team have now developed a sensor that obviates this issue by using low-depth ultrasound to image the ridges and valleys of the fingerprint's surface (and the tissue beneath it) in 3D. Though their device is inspired by sophisticated medical equipment, the scanner is reportedly very compact and only requires 1.8 V to function, making it a good candidate for use in all sorts of portable electronics.
The technology started to come together in 2007, when the researchers developed arrays of piezoelectric-micromachined ultrasonic transducers (PMUTs) which would later on turn out to be a good fit for fingerprint sensing.
To fabricate its imager, the group embedded the PMUT arrays inside a chip, and integrated it along with the same kind of microelectromechanical systems (MEMS) that are already being used in today's smartphones to build effective microphones, gyroscopes and accelerometers.
The chip, Horsley explains, is made from two wafers, a MEMS containing the ultrasound portion and a second circuit that takes care of processing the signal. The wafers are bonded together, and the MEMS portion is partially shaved off to expose the ultrasound transducers.
"Ultrasound images are collected in the same way that medical ultrasound is conducted," says Horsley. "Transducers on the chip's surface emit a pulse of ultrasound, and these same transducers receive echoes returning from the ridges and valleys of your fingerprint's surface."
Scanning a fingerprint in such a way is expected to be a more secure mechanism, and should be much more challenging to get around – a factor which will assume increasing importance as we transition to mobile payments.
According to Horsley, the use of well known, high volume manufacturing techniques means that his team's sensor could be produced at a very low cost. The hope for the future is that, besides better fingerprint scanning, the technology could also find use in personal health monitoring, or perhaps even as a low-cost ultrasound medical diagnostics tool.
The study appears in the latest issue of the journal Applied Physics Letters.