Sensors enable the smart home
05 March 2018
Real-time sensing will help define the smart home – to allow the Internet of Things (IoT) to decide exactly what the ‘things’ need to do. Widely-deployed sensing becomes the gateway to an ‘internet of awareness’ that will feed the IoT, delivering the right services at the right time.
For the digital issue of this piece, please visit this link – or click here to register for EPDT's magazine.
This will usher in a new era of efficiency and convenience that consumers will be more than willing to pay for. And while the benefits of the so-called ‘smart home’ may be obvious to the electronics industry, they may not be quite so clear to the average consumer. The key to acceptance, adoption and even enthusiasm will be the applications and services that the smart home delivers.
Moreover, as the IoT makes its first tentative steps into our homes, there is a tendency to define the ‘smart home’ in terms of the more visibly interactive examples of the technology, such as the voice-activated speakers from Amazon and Google that are becoming commonplace. But do flashy lights or a soothing voice reading us the day’s headlines really constitute ‘smart’?
A more subtle, but ultimately powerful, starting point is along the lines of the initial Nest thermostat, here the homeowner actually spins the dial (gasp!) to adjust the temperature; and in doing so, they are also teaching the device their preferences. By collecting and analysing individual data points of time, temperature and presence, patterns begin to be recognised – and with that pattern recognition comes pattern prediction, via machine learning. In the case of the thermostat, a correct prediction is rewarded by a lack of user intervention, while wrong ones are quickly corrected by the user in real time. No after-the-fact analysis is needed. The system learns, thereby introducing, perhaps, the first level of actual ‘smarts’.
No one (hopefully) makes decisions in a vacuum: data is required. In the case of smart homes (or other smart spaces), as we consider the kinds of data the IoT can make use of, we can build the list of required sensing either from the application view, or from the enabling view (see sidebar: ‘Applications vs. enabling sensor tech’). The application view starts with visualising the use-case, then plugging in the technological elements we believe can solve it. If we want to count people, for instance, we can consider the types of sensing options available, including passive-infrared, direct/higher resolution image sensing, low-resolution image sensing, time-of-flight… or something else. The category of ‘something else’ is, of course, the tricky one, as it suggests something that has yet to be associated with that type of application.
An alternative approach would be to conceptually envision throwing sensors into the space, and in so doing, enable collective creativity to begin the process of invention from there. We have essentially witnessed that kind of approach in the smartphone market space. While there was usually at least one use-case associated with a particular type of sensor added to the mobile device, understanding every use-case was not a requirement. It was more of a “build it and they will come” approach.
Add a GPS sensor for the known map application and see what app makers come up with from here! In short order, we saw navigation applications with crowd-sourced traffic updates, discount vouchers being pushed at the user as they drive near a shopping centre, and even augmented reality games where you chase cartoon creatures on your screen. How many of us could claim to have predicted anything beyond the first one?
At first glance, one could assume that smart home sensors should simply be added to the appliance(s) that makes use of them. We might expect to see applications such as appliance temperature/humidity and open/close Hall-effect sensors in the refrigerator, ambient temperature/humidity in the HVAC thermostat, or smoke/fire in the cooker hood.
But if we think about the constant elements that make up our homes, smart or not, we may reach a better result by combining sensors in seemingly unrelated hosts.
Walls, windows, furniture, lights and building occupants are all constants in our built environment. Such occupants already serve as one type of mobile sensor platform when they walk into a space with their smartphone and/or smartwatch. By applying energy harvesting techniques, wall materials could be in-built with useful sensors such as humidity or motion detection.
Lights, whether in fixed mountings or as bedside table lamps, will be ideal platforms for a broad variety of sensing. Scattered everywhere and always connected to power, they can serve as ideal costs for HVAC temperature and humidity sensors that could move off the wall, away from the wires and batteries, while sharing the BOM cost of the networking module with the connected light.
Presence, motion, activity, identity and ambient light represent disparate functions that will usefully integrate into tomorrow’s smart lighting platform. While some of the sensing elements that enable all this are just arriving on the scene, it is important to note that we already have the most critical enabler in place. Thanks to the smartphone revolution, the cloud-based applications platform infrastructure has already been developed and deployed.
If we simply imagine the smart home to be a somewhat enlarged and even more sensor-rich version of our smartphones; that application/big data platform will drive smart home decision/response patterns in real-time and predictively, just like when our smartphone pops up a notification with the latest sports score or news story we’re interested in. Prediction is at work already.
Will we see more apps and services, and will they improve? Naturally. Improvement will drive user satisfaction – and satisfied users will turn over more data. More data brings more opportunity to define the algorithms, and increases the opportunity to monetise what is learned. The virtuous cycle for success and value feeds itself. As the IoT drives massive growth in deployed sensor counts, the added data flow can’t be dismissed as a simple challenge.
The good news is that it isn’t hard to recognise the cost savings that will stem from integrating sensors into already-smart appliances (such as lights) that will pick up a big part of the load. Since the appliance doesn’t have to work all that hard to keep up with humans, there will be plenty of time and computer power available to implement preprocessing and so-called ‘fog’ computing. I don’t need to know what the temperature is right ow, as much as I need to know when it moves more than the pre-set threshold from the target. Until then, no updates required.
While we can deliver enabling technologies and visualise some of the apps, how do we drive acceptance to the convenience-loving, hassle-hating consuming public? The answer comes by combining two fundamental elements: simplicity and value. With regard to simplicity, consider the example of networking devices. Just as we have seen TCP/IP evolve from a painful, expert-only configuration nightmare into today’s plug-and-play, self-configuring dream come true, we can expect the same from the IoT in our smart home. Anyone who has set up a ZigBee-based lighting control set has at least glimpsed the start of this.
Plug it in, open the app, push the button on the bridge (one-finger security) and ‘OK’ the addition of new lights into the network. This becomes a variant of the same “build it and they will come” philosophy discussed earlier. All the key players in networking, communications and data aggregation agree publicly on the treasure trove of opportunity that the IoT and smart home represent, and they are collectively driving the standards, consortiums and platform technologies that will assure market success.
The smart home will work and users will quickly accept the services as valuable. The key will be the deployment of sensors, which no one will really want to buy as standalone elements. What they will want to buy are platforms that contain them, such as smart appliances, innovative smart lighting and other intelligent ‘things’ that provide benefits – and where yet more sensors can be integrated. Build it, and the apps will come.
Contact Details and Archive...