One day our homes will be smart, but we have a long way to go. Here’s how we’ll get there

Right now, most people don’t want a smart home. And while about 13 percent of people have an internet-connected device that isn’t a computer, phone or tablet in their homes, according to research from Parks Associates, the marketing of the smart home for the consumer is still very far out of reach.

Last week, I gave a talk on the topic before the IEEE Internet of Things standards group on how to get the consumer to adopt the smart home. I started with a basic concept.

Today, the smart home is dumb. Why? Before I ever spent $300 on [company]Philips[/company] Hue lights for my living room I had to think about why I would want them. Fortunately, because I am a person who is well versed in tools like If This Then That ([company]IFTTT[/company]), I saw the promise in linking lighting to my web services to create an ambient information delivery system that I use to alert me to stock movements, texts from my editor and the location of my husband. I also had a seven-year-old at the time and the disposable income to decide that the dance parties alone were worth it.

Screen Shot 2014-09-26 at 10.34.40 AM

But while there may be Wi-Fi on the updated Maslov’s hierarchy of needs chart circling around the internet, there isn’t a smart home. So that’s our first problem in getting consumers to adopt connected devices and the internet of things. People must want to automate their home.

There are three things preventing this. I view them as a marketing problem, a cost problem and a communications problem about privacy and security policies. We’ll discuss this in depth at our Structure Connect event next month in San Francisco with many of the companies mentioned below, but here’s my thinking today.

Why do I want a smart home again?

Despite marketers trying to convince us that we want pictures of our latchkey children sent to us when they enter the front door after walking home from school, or that we desperately need and want to spend hundred of dollars setting up an “awake” setting for the smart home that activates when they open their eyes in the morning, buyers aren’t convinced. Instead, marketers should start with the basics.

We have to figure out the conceptual smart home. Is it automated? Is it more fun? Is it energy efficient? No one knows what to do with a smart home or why they want one. It’s why I get so many emails from people asking me what devices they should buy to start building their smart home.

Inside this conceptual and marketing problem are two other issues: smart homes are shared and they are complex. Both issues require a new way of thinking about devices and how to build them compared to what tech firms have done before. For example, the technology industry has done a great job marketing computers and phones to consumers as personal devices, but the home, while intensely personal, is also shared in most cases.

So from a marketing point of view it’s hard to figure out how to sell a highly personal, yet shared device. If all of your smart devices are triggered by your phone or apps on your phone, for example, it makes it tough for guests, children or even households where some people are on Android and others are on iPhones.

The rise of the home hub

smarthingslivingroom
As for complexity, we have to start with a basic premise. The highly-detailed programmatic solution offered by many home hub companies today doesn’t make much sense. This model is one where you try to tell your home that when geofencing associated with the parents’ cell phones is triggered by both adults leaving but the two children are home, and it’s between the hours of 3pm-8pm, the following things happen.

Real people do not think this way, and most of them don’t want to think this way. In my experience what happens is you create a few highly specific settings and then walk them back over time as you realize that when the babysitter comes, your away setting breaks things, or when guests are in your home, everything becomes a problem.

Another challenge is people don’t buy smart homes; they buy point devices. And in most cases these programmatic instructions will need to join devices outside of the home in the cloud. This adds latency, additional cost to the device manufacturer and means the programming will fail when the internet goes down or APIs break.

Today hubs are the solution, generally with enough compute power and some form of mesh or direct Wi-Fi technology to ensure that the hub can still act when the power goes down. But this is still a strategy that requires the user to think like an engineer and to plan their automation. And if they are like me, when they do this they will at first go overboard and then gradually walk back to the minimum when they realize they don’t actually always want the coffee maker to turn on when they enter a certain hallway in the morning; you might be on your way to the airport in a rush instead of your normal routine.

Let there be mesh

The more interesting conversations right now solve these issues by either imbuing the device with smarts via learning algorithms or by trying to implement some kind of communications between devices, like AllJoyn or the Open Intercommunication Consortium. In one scenario you have devices that are using contextual clues from the environment and any web services it can access to deliver what it thinks you want when it thinks you want or need it. However, so far, these algorithms tend to be opaque. Nest is a good example of this approach.

In the case of AllJoyn or some kind of standard — [company]IBM[/company] is looking at using the block chain, Telehash and BitTorrent as part of open source code called Adept for example — they are interesting but rely on some kind of universal adoption to make it worthwhile. I look at them and worry that they could be a reprise of DLNA, which was held as a panacea of audio-video compatibility but was a gift to Hollywood, and ultimately didn’t always work.

One way to ensure that devices that talk to each other and are working the way they should comes from Tom Coates of Product Club. He has written about a kind of home device chat room that a consumer could log into to see how the home was making decisions and guide it. It’s an interesting proposition, although it would require a bit more in terms of artificial intelligence to become a reality. But if I could log into that chat room the night before my mad dash to the airport and make sure the home’s butler picked up my flight from my calendar and knew what to do, that would be awesome.

At the end of the day I want to be a manager of my home, not a micro manager programming every decision my home needs to make. Yet, I can see what’s happening and influence how the home decides.

The smart home is too expensive!

Connected devices can cost five or more times as much as the devices they are replacing. For example, the Nest thermostat at $250 replaced a Honeywell that was about $70 in my home. The forthcoming August lock at $250 costs way more than a typical $30 Kwikset lock. Between lightbulbs that cost between $30 and $60 a pop and connected blinds, locks and more, I’ve replaced a dozen or so sub-$100 or even sub-$50 products with gear totaling a couple thousand dollars.
Screen Shot 2014-09-26 at 10.35.16 AM
Most people aren’t as enamored of tech as all that, especially given the marketing problem discussed above. There are of course ways to solve the cost issue. One way is to sell a status device. Again, the Nest is a good example.

I once was at a dinner party at the home of a friend who is the CEO of a company outside of the tech industry. I noticed the Nest on her wall and asked her about it. She told me that she loved the device and it was a joy to touch. In fact she had come home early that day to adjust the temperature down for the party and didn’t even mind because she liked playing with the device.

She came home to turn down the Nest. She drove home to change the temperature on a device that can be controlled from her phone. This was not a dumb woman, but clearly she didn’t care about the remote access feature of the thermostat. She don’t realize it existed. The Nest was a beautiful object in her carefully appointed home. She also valued the learning aspects of the device.

And those learning aspects bring to another way to bypass the cost problem. You change the conversation; the [company]Nest[/company] isn’t a thermostat, it’s a learning thermostat. Remote access isn’t a highly marketed feature. Obviously you pay more for something that learns.

There is a similar conversation around [company]August[/company] locks. CEO Jason Johnson will tell you this isn’t a lock, it’s a robot. This is a robot that follows your instructions and allows you let people into your house even when you aren’t there. Will that offset the $250 cost of the lock? No idea.

The cost of connected devices will have to drop. We are at the early days of this shift in manufacturing consumer products and the cost of the Wi-Fi modules and microcontrollers will gradually fall as demand increases. Moore’s Law will see to that. But most consumers don’t understand the cost profiles of different radios nor do they understand that the more industrial devices such as those capable of mechanical action at the flick of your phone, require precision motors and parts that can greatly add to the price tag.

Existential worries and legit ones

Screen Shot 2014-09-26 at 10.35.44 AM
The final issue in the smart home can be found in articles about viruses invading your light bulbs or police having access to your IP camera footage on Google’s servers. There are many models and scussions taking place on these topics, from a 5-point plan for handling security in a smart home or car to efforts to use block chain technology to enforce privacy across many different devices and locales.

But I think the consumer will have to play a role in updating their passwords, patching their systems and keeping abreast of potential issues. The device makers will have to do the same, and take it one step further with better labeling for security updates and more attention to continuously testing for security flaws. When something is connected to the internet, its chances of being compromised don’t stop the minute it’s shipped.

The government has a role as well. On the security side the FTC should fine companies who have repeated issues or don’t update their software quickly enough. On the privacy side, it needs to come up with rules around metadata that go beyond just preventing companies from sharing Personally Identifiable Information. We also need to extend civil liberties against unreasonable search and seizure to digital devices and artifacts.

The bottom line

As you can see, handling these issues isn’t as simple as telling a consumer they can spend five times the cost of a normal device and get a connected experience. A salesperson at [company]Savant[/company], a high-end home automation provider, once told me that in a few years people will view home automation like they view indoor plumbing. Sure, this is the man selling those connected toilet, but when my home works perfectly, as sometimes happens, I think he’s right.

The key is giving consumers a reason to buy a connected device by addressing the security and privacy issues as a community, while also understanding the cost and features of your device and how that will need to play into your design and marketing.

We will need to tackle legislative, design, engineering and communications issues and then translating that back into services and products the consumer can use without having to meticulously plan and orchestrate their lives. After all, we are humans. We build tools not so we can adapt to them, but they can adapt to us.