Smart Cards with Built-In Fingerprint Scanners

A startup plans to market cheap scanners that use printed organic light emitters, detectors, and electronics.

Paper-thin fingerprint scanners could soon make cell phones, USB memory, and credit cards more secure. The new scanners, developed by Nanoident Technologies, a startup based in Linz, Austria, detect patterns and blood content in the tissue within the finger. This makes them much harder to fool than ordinary fingerprint readers, which measure only the ridges and contours of fingerprints.

Nanoident's technology, which could be available in consumer devices within two years, takes advantage of inkjet technology to print light sensors, light emitters (similar to the organic light-emitting devices used in some cell phone displays), and electronic circuitry for memory and signal processing. This combination makes it possible to illuminate an object and measure the reflected light, says CEO Klaus Schroeter.

The printing-based manufacturing process is far cheaper than the process used with silicon-based electronics -- the sensors will cost less than two dollars in cell phones, and in large-volume "smart card" applications, less than a dollar, Schroeter says. Furthermore, the devices can be printed on a wide variety of surfaces, such as glass, plastic, and paper.

The scanners combine information from the surface of the finger -- the fingerprint -- and the underlying tissue and blood. This deeper information is obtained by illuminating the finger with different wavelengths of light, some reflecting off the surface, and others, toward the red end of the color spectrum, penetrating deeper, allowing the scanner to extract data from beneath the skin. Combining information about the tissue structure, which Schroeter says is unique to each person, with conventional fingerprint data improves the accuracy of the scanner. The company’s sensors are more than 99 percent accurate, Schroeter says, while conventional fingerprint detectors are inaccurate for around three percent of the population.

The more detailed measurements make the sensors difficult to fool. According to Peter Honeyman, scientific director of the Center for Information Technology and Integration at the University of Michigan, making artificial fingers to trick conventional sensors is easy -- and an increasing threat to fingerprint-based security. But since such a fake finger does not have the tissue and blood of a real finger, Nanoident’s sensor could tell the difference.