Fingerprint Technology or Biometrics?
Contributed by Martin Bijman
Apple has raised interest in biometrics with the launch of the iPhone 5s and its new Touch ID feature. Biometrics stock valuations jumped briefly around the launch date as the market took a closer look. After all, biometrics is a good marketing term for Apple to try and adopt – biometrics may be the basis for many future Apple products.
In this blog, we consider whether Touch ID is in fact biometric, and compare the relative patent landscapes of fingerprint and biometric technologies. We also take a look at who Apple may have to watch out for in terms of early market adopters.
Fingerprint authentication has been around for years, traditionally sliding a fingerprint across a 1D scanner. These products have generally been viewed as functional but unreliable. Apple’s Touch ID may be the highest volume product that uses a 2D fingerprint imager. It’s fast, it’s working well, and is more convenient than a PIN or slidebar. But is it secure? Probably secure enough for a consumer product, but not as good as a PIN. Curiously, clever people are posting videos of lifting fingerprints off the dirty touch screen and creating a 3D relief that can subsequently unlock the phone.
Is Touch ID really biometric? According to Wiki, biometric “refers to the identification of humans by their characteristics or traits.” This implies processing human sensor inputs of many types, including fingerprint, and comparing them to stored templates. Applying this definition, Apple’s Touch ID would fit into the biometric category.
We also took a sample look at the patent landscapes of biometric and fingerprint technologies. Using about 1200 patents, we took a look at the split by priority year. See chart 1. To our surprise, the patent landscapes essentially do not overlap – only about 10 patents are shared.
However, taking a closer look at the landscapes through keyword analysis illustrated something quite different. See chart 2. Here we see the landscapes beginning to blend, with biometric (blue) and fingerprint (red) overlapping in some areas, such as processing. Fingerprint patents appear to be more about sensors and partial scans. Fingerprint inventions were earlier, it’s a mature and deployed technology. Biometric innovation is also rich, but is predominantly about authentication. Biometric deployments have been modest, limited to government and emerging areas like face and voice recognition.
From a market perspective, who are the early adopters of these technologies that Apple may have to watch out for?
Apple is in a good position in the fingerprint patent landscape. The fingerprint landscape suggests the best holders of early patents are NEC (17 patents up to the year 2000) and Fujitsu (13), also Activcard Ireland and Fujifilm each hold a couple early patents. Apple (21) and Sony (12) hold the next best positions with patents as early as 1996. Apple’s position is based on patents from Authentic and Harris Corporation.
In contrast, the biometric patent landscape suggests best holders of early patents are Activcard Ireland (46 patents up to the year 2002) and Open Invention Network (14), followed by Omega Patents LLC (5) and Lasercard (2). The best of the rest include IBM, HP, Imprivata, and Fujitsu. Apple appears to hold three early patents of priority 1997 from Authentic on “expandably storing and searching biometric data,” then the next block of 10 patents is for priority year 2003.
Of course, searching for just the term “biometric” is limiting. If one looks more generally for patents about recognizing or authenticating human traits such as eyes, voice, fingers, handwriting, and heart, 136 Apple patents were found, with a notable increase in the last five years; see chart 3.
In summary, indeed the Touch ID feature is a biometric authentication of a fingerprint image. By grabbing the marketing word biometric, Apple is setting up the future of its product lines to input and process many types of human traits and behaviors. Their patent position for both fingerprint and biometrics appears to be good, considering this initial study found early relevant patents in both markets.
Other Articles of interest: