The Need For Security in IoT
Increased connectivity is leading to all sorts of technological innovation, bearing the potential for a great deal of increased efficiency, not only in the world of industry, but also in at home and in the car. The Internet of Things, both industrial and commercial, promises levels of automation, customization, and convenience that stand to improve our relationships with our environment. Our door locks, our thermostats, our smoke detectors, our fridges and light bulbs; our work terminals, our imaging machines and transportation systems; our children’s toys, our watches, our televisions, our washing machines and baby monitor cams—all of these and more are connecting to the Internet to make our climate more comfortable, our day more efficient, and our energy consumption more responsible. But with every connection to the Internet there is also a risk of a hack attack.
The Internet of Things is vulnerable to hackers. Experts have been espousing the dangers of an unprotected IoT for years now, and we have already seen the results. Automobiles have been hacked to demonstrate this vulnerability, and there is a web app that enables anyone with an Internet browser to spy through unsecured connected cameras in people’s homes. Even sex toys can be hacked, according to Motherboard’s Nicole Kobie.
Whenever the issue of an unsecured IoT is brought into the light, it is inevitable that biometrics are floated as a security solution. Appealing for their convenience and ability to mitigate the scale of hack attacks, biometric security solutions offer an answer to the question of unsecured machine to machine networks. What’s more: biometrics are appealing to consumers, as interested smart home buyers have expressed concern with the idea of having to potentially secure smart devices with PINs and passwords.
Of course, multimodality is the name of the game in the vertical markets these days, but when it comes to IoT, one type of biometrics is king: voice recognition.
A Paradigm of Convenience
As a contactless modality, voice recognition syncs up perfectly with the Internet of Things paradigm of convenience and efficiency. It is a strong authentication solution that is mature enough for a wide range of deployments and is software based, requiring only a microphone on a device to act as a sensor. Voice biometrics tech is invisible and can allow for authentication at a distance, so verifying your identity to a smoke detector won’t require you to grab a chair or step ladder.
Admittedly, facial recognition offers much of the same convenience of voice biometrics, but where the spoken modality really makes its gains is in concert with its sister technology: speech command. Speech recognition, which allows users to speak to a connected machine in order to interface with it, is a natural fit for IoT devices, as many do not have display interfaces. In concert with voice recognition, devices can discriminate between users, understand which commands ought to be listened to and which should be ignored. It’s an advanced sounding modality, but such technology is already widely available and in use, as Apple’s AI assistant Siri was upgraded to offer this functionality late last year.
Talking To Your Things
The Internet of Things industry seems to be moving forward in a way that is agreeable to the assertion that voice biometrics are the modality of choice in the hyper-connected world. Already we are seeing IoT platforms and chipsets ship with integrated voice biometrics technology. Samsung ARTIK chips support Sensory’s TrulyHandsfree speech recognition, as does Intel’s system-on chip line of products. In the automotive market, hands-free command is becoming a standard feature too, albeit primarily as a user interface option, though airplanes with hands-free tech can distinguish between voices.
Indeed, it is because voice biometrics are so easy to deploy on small devices, and integrate so well with hands-free interface technology, that they should be the prime modality in the Internet of Things as it starts to lock itself down in anticipation of mainstream market growth. We can already see voice recognition gaining traction in this next generation mobile market, with wearables and automotive markets particularly ripe for adoption right now. All that remains to be seen is if other biometric types will manage to catch up to the head start voice has benefited from thanks to its ability to so thoroughly fill the M2M niche.
*
Stay posted to FindBiometrics throughout May as we continue to examine the featured topic of voice biometrics. Participate in the conversation by following us on Twitter.
Voice Biometric Month 2016 is made possible by our sponsors AGNITiO and SpeechPro.
—
May 11, 2016 – by Peter B. Counter
Follow Us