“Audio analytics” might sound like a geeky, newfangled concept, but the technology has been around for quite some time in the security industry, where glass-break detectors are calibrated specifically to detect the sound of breaking glass while ignoring similar sounds such as coins dropping on a marble countertop.
Similarly, we have seen “video analytics” at play in security cameras for a very long time, starting with basic motion detection: When pixels are disturbed, we know something in the space has changed, so wake up, security guards, and look at the monitor.
Audio and video analytics have come a long way, especially in the past two years and especially in the consumer market. Cameras (with their on-board or cloud-based processors) can now recognize faces, detect fires early, count people, detect fever, and even catch the early onset of dementia.
Listening devices can detect water leaks and termites behind the wall, distinguish between cries of anguish and joy, and issue alerts based on the sound of doorbells, smoke alarms and microwave dings. Following is a look at how these audio and video analytics trends are coming to market.
Audio Analytics: The Detection Specialists
The Consumer Electronics Show (CES) in 2015 marked the beginning of mass-market audio analytics for the smart home. At least a dozen companies announced new Internet-connected products that would listen for the blares of “dumb” smoke/CO detectors, and then alert the homeowner via email or text message. CentraLite, First Alert, Kidde, Leeo, Roost, Swann, Wemo and others all introduced these types of products, usually retailing for $100 or less.
Going further, newcomer Cocoon introduced a consumer product featuring its own “Subsound” technology, which learns the sound patterns of a home, and alerts homeowners when unusual noise is detected — not just gun shots, for example, but wood creaking or water dripping behind the walls. The company utilizes infrasonics — the same low-frequency sound monitoring used by seismologists — to capture inaudible activity throughout the premises.
Listnr is (was) another consumer-electronics maker working on listening technology, although its project was suspended on Kickstarter in 2015. Even so, the initiative reveals the potential of audio analytics. Listnr could be programmed to respond to various sound commands such as snapping, clapping or foot-stomping. It purported to discern between crying, laughing and screaming. This is all technically doable, just not by Listnr.
Many of the real listening devices feature intelligent sound detection technology from U.K.-based Audio Analytic. Cisco and Next Level Security, for example, use the technology for their IP cameras to detect car alarms (from seven standard suppliers), gunshots (and their unique muzzle blasts) and aggression (measuring acoustic changes in voice). Louroe Electronics, a leader in audio monitoring and verification for security and productivity applications, recently added the Audio Analytic technology in its enterprise-grade microphones and processors.
On the consumer side, Sengled recently introduced a smart bulb called Sengled Voice that incorporates a microphone for both voice control of smart-home systems, as well as audio analytics. The company says the bulb provides “enhanced security by detecting sounds like glass breaking and babies crying.”
Video Analytics: Expanding ID Potential
Security agencies have used extensive video analytics for years, for example, to distinguish between people and animals at an illegal border crossing, identify and track objects in an airport, and verify identities through facial recognition. Commercial enterprises might use the technology to analyze traffic flow at a store, estimate the demographics of guests, or measure worker productivity.
One of the governing standards for video analytics, OpenCV (computer vision), has been exploited to extend analytics to many new applications, including such things as flame and smoke detection, enabling advanced warnings before traditional sensors detect heat and smoke-related chemicals.
Bosch demonstrated in 2015 a camera that picks up a small flame in a big open warehouse long before traditional sensors detect the danger.
Now, this technology is being implemented in consumer-grade products, primarily for facial recognition, with more to come. It used to be that a camera would capture motion at the front door, and then start recording, possibly alerting the home owner with a five-second video clip. Now these cameras can determine who is at the door, so the system can save storage space and preserve battery life by recording only when an unknown guest is knocking.
Between Intel’s new RealSense camera technology and implementations like Microsoft’s Windows Hello, we can see how facial recognition might become the de facto login mechanism for computers, phones and other devices. We can also imagine how an automation touchscreen could display your personal preferences as you approach the monitor because it recognizes your face.
At CES 2016, several buzz-worthy consumer cameras like the Netatmo Welcome, Simplicam by Closeli (powered by ArcSoft facial recognition), and Sengled Snap, a screw-in outdoor LED floodlight with a built-in camera. The new Tend Baby app, powered by Kodak, does all the usual baby-monitoring stuff, but also includes algorithms to track the sleep patterns of the cribbed one, and to alert parents when their little one is “awake and agitated.”
More Implications in the Home
Soon, the right combination of hardware, software and algorithms will be used to replace many of the dedicated sensors used in homes today, from motion sensors to flood detectors.
But they can do so much more, bringing “smart” homes to a whole new level.
For example, a Spanish technology firm called Tecnalia is working on a project for early detection of dementia, using sensor networks in the home and some sophisticated processing. The system would go into a house while the resident is relatively healthy, in order to establish a “normal routine.” Over time, sensors would pick up changes in activity such as slowing down, skipping meals, restlessness and certain signs of Alzheimer’s — such as when a person stops in his or her tracks and reverses direction, possibly signifying disorientation.
Julie Jacobson, recipient of the 2014 CEA TechHome Leadership Award, is co-founder of EH Publishing, producer of CE Pro, Electronic House, Commercial Integrator, Security Sales and other leading technology publications. She currently spends most of her time writing for CE Pro in the areas of home automation, security, networked A/V and the business of home systems integration. Julie majored in Economics at the University of Michigan, spent a year abroad at Cambridge University, earned an MBA from the University of Texas at Austin, and has never taken a journalism class in her life. She’s a washed-up Ultimate Frisbee player currently residing in Carlsbad, Calif.