There’s a lot of talk around the Internet of Things these days, but one trend we’ve noticed involves merging several traditionally separate products into one hybrid Smart Object that communicates with mobile devices. Further, many of these devices utilize the power of sound. They listen to the ambient sounds around them and behave accordingly. Unlike Siri, Cortana or the Amazon Echo, these devices are more specialized, designed to listen only for a single, specific sound (or for a couple of similar sounds). As Roost CEO Roel Peeters says, “Who really cares if the dishwasher sends a message to my phone telling me the dishes are clean? Instead of connecting devices, we should focus on making data connected.” Well, sort of. But semantic-accuracy aside, the premise still holds serious promise.
It started with the Roost Smart Battery. Introduced October 2014, the Roost is a WiFi 9-volt battery that powers your smoke detector and alerts you whenever it goes off. The Battery works by listening. When the smoke detector goes off, it ‘hears’ it and alerts whoever’s been designated to receive the notifications (so long as they have the Roost App installed—a harbinger of more UX issues to come). The Roost also uses sound (from your Smartphone) to setup its WiFi connection. Tones emit from your phone’s speaker and seconds later the battery is connected.
Amazon reviews are mixed and with good reason. The concept is misguided. Most users complained about the battery not being powerful enough to power the smoke detector, or worse yet, going “offline” without giving anyone a warning. UX Rule #1 – ALWAYS provide offline notification along with an offline experience. Obviously, a user whose smoke detector battery loses connectivity without warning or notice is far worse off than a user who only expects his smoke detector to chirp when the battery runs low.
IoT Rule #1: Don’t use new, unproven technology to help people keep their homes from burning down.
IoT Rule #2: Don’t then entrust this new, unproven mobile technology that keeps people’s homes from burning down to a battery.
A week later, a company called Leeo had a better idea. Instead of putting the smart listening device in a battery, why not plug it into the wall? Why not add a night light? And why not email users a recording of the trigger-sound—a brilliant way to let a human verify or reject the alarm, especially since the device probably records the sound anyway in order to analyze it. Enter, the Leeo Smart Alert Nightlight.
Then in mid-2015 came the Twist, a light bulb that also contains an Airplay-enabled speaker. It doesn’t listen for anything, but it does make sounds and it is a hybrid Smart Object so it kind of applies. Also, the clever thing about putting smart hybrid technology in light bulbs (as opposed to inside batteries) is the bulb socket can be used as an electrical outlet, powering whatever device is inside the lightbulb—in addition to the light bulb itself.
In late 2105, the BeOn Bulb took things one step further and created a similar hybrid smart bulb, only this time they stuck a battery inside a light bulb! Why would you need a battery in a light bulb? I’ll explain. The battery basically turns the bulb into a flashlight. What’s more, the battery is charged by the light bulb socket such that when the power goes out…the lamp will still light up! Voila. Instant emergency lighting (it smartly uses bluetooth so it works when a power-outage kills your router). Also, users can still turn the light on and off from their smartphone even if the lamp is switched off. Better still, this thing also contains a microphone that listens for your smoke/CO alarm, doorbell or burglar alarm and turns the light on when it hears them. It’s a security system in a light bulb!
But for one simple reason all these products fail to live up to their full potential. They’re all missing the one feature that could be used to turn practically anything into a smart object—the ability to tell the device what sound to listen for. All users would need to do is first sample an object’s sound, then use the sample as a reference point, an audio signature. Since everything has its own unique audio signature, leveraging this signature should be as easy as sampling the sound a few times and using the average as the ‘key.’ The device could even store multiple samples. One, say for the doorbell, one for the front door opening, another of a smoke alarm going off, another of the baby crying. Each sound would be assigned a unique label to identify it along with a specific set of instructions to carry out when that sound is heard (email grandma, send SMS to X and Y, dial 9-1-1, etc.).
The point is, this technology could be used to detect anything. Anything, that is, that makes a sound. Put the device in listening mode then open your front door and suddenly you’ve got a device that will alert you whenever it hears your front door opening. Create a more accurate baby monitor that alerts the user only when the baby is crying. Get notified whenever your heater kicks in. The sky’s the limit!
If your company would like to discuss ways it might streamline business processes or improve employee satisfaction by incorporating the IoT into its enterprise mobile strategy, give us a call. We’d love to talk.
Content Strategy Lead at Anexinet