What I Learned About the Smart Home From Building a Home

We just finished having a new home built and moved in this past weekend. I’m a technologist by trade and have always loved gadgets. With a green field site, high tech awareness and a love for gadgets, you’d think my new home would be filled with home automation and smart home technology. Yet there’s remarkably little of this technology in our new house. Why not? I’ll talk through the two pieces of this technology we have installed as a way of illustrating why I think the smart home still has so far to go.

AT&T Digital Life – a complex installation

Earlier this week, we had a technician with an AT&T uniform and truck (technically, an employee of a third party contractor AT&T works with) come and install the necessary equipment for AT&T’s Digital Life service in our home. He arrived at 1pm or so and left around 6pm. What was he installing? A simple security system, with sensors on several doors, and a garage door opener.

Why did it take so long? Well, here’s the problem with almost all new technology in the home: it’s very much tied to old technology. I’ve had fiber broadband connections installed in three homes now and, in all three cases it took about half a day or more to install, as the technician first installed a box somewhere on the outside of the house and then installed a box and ran cabling inside the house. With our Digital Life service, connections had to be run to various power outlets and our broadband router and various boxes installed inside the house. Then the sensors had to be installed on the various doors. Then these had to be configured and tested. Then the technician had to install the garage door opener. All this for a system that actually runs off cellular signals.

After all that, what do we have? A smart, connected security system that can monitor when various doors are open and closed, set off alarms as appropriate, and so on. And a garage door opener that – for now at least – doesn’t work (apparently a compatibility issue with our brand new garage door mechanism that will hopefully be fixed soon).

I’m not complaining about any of this – I chose to install it and was made aware how long it would take. But it illustrates the challenge of installing new technology even in a brand new home. This stuff is hard, it’s disruptive, and it takes significant time even for professional installers to put in place.

Nest – simple, but perhaps too simple?

The other piece of technology we have in our home is two Nest Learning Thermostats to control our two zones (upstairs bedrooms vs. ground floor and basement living areas). I love the Nest – it’s easy to use, looks beautiful and, because it was installed as part of the construction, it was easy to install (though our builders did that job for us).

However, we could potentially have seven zones in our house – the HVAC was set up that way – and we’re finding we probably need those seven. The different rooms in the house are hit by the sun at different times of day and tend to heat and cool at very different rates and at different times as a result. So just because the Nest is in one room in a particular zone doesn’t mean the temperature it senses there is the appropriate one for the other rooms in that zone. We really need all seven of those zones to keep the temperature comfortable in all the various rooms. But that means installing not just two but seven of these $250 devices. While one can justify spending $500 on two of these on the basis of cost savings from the smarter HVAC control, spending over $1500 is a lot harder to justify. Yet these devices cost vastly more than the basic thermostats our contractor would have chosen by default.

Cost, complexity and interoperability issues

These two experiences highlight some of the major issues the smart home faces today. High cost, complexity of installation and usage, and interoperability issues between devices. Nest is creating its own smart home ecosystem, and AT&T’s Digital Life is an umbrella brand for its version of a smart home platform. Neither talks to the other and neither is part of the Apple HomeKit ecosystem or other major consortia. So, even though I might, in theory, want to be able to turn off the HVAC and arm the alarm with one simple “I’m leaving the house” gesture, in reality I have to manually set both of those things in separate apps (or by hand) if I want them to occur as I leave the house.

In addition, the AT&T experience in particular highlights what I think is one of the key challenges with the smart home: so many of the things we want to automate in our homes exist at a very granular level. Just as every door we wanted to incorporate into our alarm system had to be given its own sensor, so every light switch (or light bulb) would have to be controlled independently if we wanted to automate or remotely control our lighting with any precision. Thermostats are ripe for disruption partly because there are relatively few of them in the average home – our two are probably pretty typical. But light switches, electrical outlets, blinds, doors and windows are in the dozens in many houses today and it simply isn’t practical either to install smart home solutions on all of them or to pay the high cost to do so. I’m not sure I see a short term solution to any of these challenges, even though the interoperability issues will likely be worked out in time. The sheer complexity and cost is what makes smart home deployments prohibitively difficult today, despite the appeal to a gadget lover like me, and I don’t see that changing anytime soon.

The Quantified Car

It is easy to make the assumption that most if not all electric products will have a companion app or be managed by a software platform in the future by a smart screen of some kind. Trying out these solutions as they stand today help to make sense of what they may look like or how they may be used tomorrow. Lately, I’ve been trying a product from a company called Automatic which links to a software app and begins tracking important data related to a car. I think of it like a health and fitness wearable for your car.

You plug Automatic’s hardware into your cars diagnostic port. It then connects to your smartphone with Bluetooth 4.0 and begins tracking things like time, distance, MPG, and how much money in gas you spent during each drive. At the end of the day or week you have a running tally of all those stats for the day or week.

As I started using this it felt like my first experience using a health and fitness wearable. Until such devices came along we had no real way to consistently measure things like steps, time spent sleeping, heart rate, and more throughout our days. Getting all this data was eye opening to me. Then the question becomes what do you do with it?

In the case of health and fitness wearables, tracking steps is useful if you have a goal of staying active or getting a certain amount of steps in each day. The data is useful in so much as you use it to modify your behavior in ways you see fit. And so the folks at Automatic added similar features into their solution. Namely, getting better gas mileage and decreasing wear and tear to your car.

This is accomplished by audible sounds, in this case a beep, when you go over 70 mph or when you accelerate too fast from a stop. Both driving over 70 mph for long periods of time and accelerating too fast are known to yield poor gas mileage. The other sound it makes is when you brake to hard. This increases wear to your car’s brakes and overall safety and is recommended to avoid. While I have no way to quantify the braking component in terms of tangible results, I did notice that my car’s average MPG went from 32.1 MPG to 37.7 (I have a Kia Optima Hybrid) just by abiding by Automatic’s audible noises which alert me when I went over 70 mph and decreased my speed to less than that. Another case of using data for quantified results which I would not have had without the help of connected hardware and companion software.

All of this brings up an important point. The key to quantified hardware is in its customization. Everyone’s goals and habits will be different. So the software needs to be dynamic enough to allow a full range of customizations for the person who is using it. These types of solutions, whether they be health and fitness wearables, connected car, connected home, connected watch, etc., all need to be highly customizable for the specialized interests of the buyer. This is why a one-size-fits all solution will not work in this space.

While Automatic’s solution is interesting, I assume that this type of technology will be built into all cars in the future. Your car will come with an app, your appliances will come with an app, even your toilet will come with an app that monitors their use in some way. Diverse connected hardware offerings combined with robust and dynamic software is what will turn these visions into reality and bring them from the early adopters to the mass market.

Why We Keep Hearing About “Internet of Things”

I must admit I’m really excited about all the dialogue surrounding the Internet of Things. I’ve spent the last eight years focusing on the Connected Home, and now I’m finally seeing some of the use cases from my old PowerPoints becoming a reality. Analysts and startups are now using “Internet of Things,” or IoT, to describe the hyper-connected world we’re predicted to inhabit within the next few years. But it’s not only the startups… GE, for example, recently unveiled Project Wink in conjunction with innovation community platform Quirky making thousands of patents available for the creation of connected devices for fleet management, healthcare, and sustainability. Cisco is heavily marketing their “Internet of Everything” vision complete with multipart whitepapers, abundant data sets, and even a resident futurist. Even Intel released a blog post this spring celebrating International Internet of Things Day on April 9th.

So we’re seeing a broad usage of the term “Internet of Things,” but what does it mean? And what are the underlying components setting the stage for IoT to explode. In this post I’ll break it down, review the reasons for its rapid growth, and most importantly propose why it matters to us as technologists.

What is Internet of Things?

Let’s first define “Internet of Things.” The Wikipedia definition is a good start: “The Internet of Things refers to uniquely identifiable objects and their virtual representations in an Internet-like structure.”

So basically imagine a variety of our physical devices—such as lamps, refrigerators, cars, watches, etc.— having a fully digital representation complete with accessibility and an address. We are only beginning to understand the implications of such widely available data, and the impacts to commerce, privacy, and consumer behavior are all highly debatable (will save for a follow-up article), but here’s some industry data on the relative scale of the IoT sector:

  1. By 2020 a cumulative 100 billion processors will have been shipped, each capable of processing information and communicating. (Source: Ericsson)
  2. Only 0.6% of physical objects that may one day be considered part of Internet of Things are currently connected. Between 2013 and 2022, $14.4 trillion of value (net profit) will be “up for grabs” for enterprises globally. (Source: Cisco)
  3. Across the health-care applications analyzed, Internet of Things technology could have an economic impact of $1.1 trillion to $2.5 trillion per year by 2025. (Source: McKinsey)
  4. 15% of surveyed organizations across the globe already have an Internet of Things solution in place, 53% plan to implement one within the next 24 months, and another 14% in the next two to five years. (Source: Forrester)

Examples of Internet of Things

Many of the devices and services we use on a daily basis have been put into the category of IoT. Here are some examples:

  1. “Smart” devices such as refrigerators, electric meters, doorlocks, etc. that use connectivity and intelligence. Nest and Silver Spring Networks are notable in this category gaining a lot of attention and interest from the channel, potential partners, and customers.
  2. Wearable computing for health & fitness such as the FitBit, Jawbone Up, and Basis band as well as Independent Living platforms such as Lively and BeClose.
  3. Industrial-grade devices and applications that aim to improve retail experiences (Euclid, RetailNext), energy efficiency (Enlightened, Greenwave Reality), and healthcare delivery (Telecare, AliveCor).

And these really aren’t exotic use cases that need to happen to make a material impact. For example, home appliances that can identify maintenance required and trigger a proactive service call, a smart meter detecting presence in the home to optimize grid utilization, or even location services like Foursquare sending succinct information to a smart watch.

Matt Turck and the team at FirstMark Capital have done an excellent job of listing some of the major players in the graphic below. There are more companies operating in stealth mode or outside of the traditional VC ecosystem (i.e. inside a research lab at a public company or university), but this offers valuable insight into the various components.

Screen Shot 2013-06-02 at 9.07.12 PM

Reasons for the Internet of Things Momentum

As previously noted in analyst reports, the term “Internet of Things” was coined backed in 1999. But when looking at the overall activity in this sector, things started to inflect around 2011 with the visibility of Quantified Self products such as FitBit and Nike Fuel.

Prior to 2011, my CES experience as an exhibitor in the South Hall between 2007 and 2010 saw incremental improvements to both floor traffic (read: reporters, prospects, and VCs) and interest levels. However, the past two years the South Hall was abuzz with a variety of stakeholders looking to stop by and see what was happening at the ANT+, BodyMedia, Dropcam, FitBit, Nexia, Polar, etc. booths and the respective Z-Wave and Zigbee alliance pavilions. This year saw a lot more coverage, tighter focus on creating end-user value, and exponential growth in interest.

Looking beyond just the hype factor, there are a variety of reasons for the acceleration of Internet of Things. Most notably, but not limited to, the following:

  1. Chip companies via Moore’s Law continue to make smaller, faster, and lower-cost silicon that can embed in new form factors and provide high-efficiency processing.
  2. Networking progress including more down/up stream bandwidth, more reliable Wi-Fi, NFC, Bluetooth LE connections, and hockey stick adoption of smartphones.
  3. IPV6 rollout increases the number of available “internet addresses” by a massive factor. IPV4 could handle roughly 4 billion addresses whereas IPV6 can handle 340 trillion addresses.
  4. Big data tools and methodologies now available to process the magnitude of information coming from connected devices ((caveat: insights and actions from the data still lacking. I suggest an excellent presentation from Greylock’s DJ Patil here:))
  5. Lower barriers to entry with open source hardware options like Arduino and Raspberry Pi, software from Amazon Web Services, and crowdfunding options including Kickstarter.

So Why Do I Keep Hearing About It? Is the Time Now?

In a word, yes. This time is now for staging the opportunity. At 4Home (and most startups that pursued Home Automation in the 2000’s) we realized that we were early to market. There was a sector of hobbyists looking to leave X-10 and a smaller support network of insiders in press, venture, and corporate partners who could push the momentum along. As we were acquired by Motorola and later became a corporate partner in the ecosystem we began to see the market conditions improve. It was no longer Jetson’s inspired serial entrepreneurs; it was a new group of technologists focused on larger and more ambitious projects. Specifically:

  1. Focus on new products that deliver direct end-user value via B2C business. Nest and Dropcam are two examples of companies that have a clear value proposition and execute through a hybrid retail/direct model.
  2. Trade press, analysts, and influencers writing about the sector and people are talking about it more. I’ll make a causal assumption the increased coverage is indicative of increased demand.
  3. Venture Capitalists, who largely stayed out of funding Home Automation, are being active in developing/managing IoT theses. Some notables I’ve met include Tim Chang (Mayfield Fund), Trae Vassallo (Kleiner Perkins), Rob Coneybeer (Shasta Ventures), Mike Dauber (Battery Ventures), and Jason Krikorian (DCM). Others are working closely with Venture Partners or EIRs to atomize the sector and develop investment options.
  4. CRITICAL – talented engineers, designers, product mangers, etc. look at this sector to start or join companies. The flow of high quality engineering talent from Apple, Facebook, Google, Intel, etc. into these companies like SmartThings, Electric Imp, or Jawbone is fundamentally the reason why I’m so excited about this space. Great products need strong benches and it’s great to see a sector I’m very passionate about finally get the attention it deserves.

So while these factors signify a discernible improvement in market conditions the ultimate (and largely unproven) test of IoT will be the wide adoption of these solutions at scale with consumers and in the enterprise. And consequently, do we see a vertical ecosystems develop without wide interoperability or will broad based standards dominate to serve a common open environment? I’m thinking very heavily about this.


The blatant overuse of terms like “smart ____” and “connected ____” have made some of us (rightly) skeptical about innovations like Internet of Things. But in the case of Internet of Things we will continue to see gradual changes in the near-term (next 12 – 24 months) leading to significant, longer-term shifts in how consumers go about their daily lives and how business is conducted around the world (36+ months). Get ready for one of the most interesting, if not most disruptive periods, in modern tech history as this plays out in your home, at your office, and in the network.

I’ll Take My Hardware With a Side of Software

samsung-fridge-2Despite some people loathing CES, I actually happened to enjoy the show this year. I go to CES mostly to meet with our clients, gather data and market intelligence, and search for trends. CES always includes hidden gems; you simply have to know where to look.

Two things stood out to me as major themes at CES this year.

The Internet of Things on Display

We have talked about the concept of the Internet of Things for several years now. The Internet of Things is the idea that the vast majority of our electronics will be connected to the Internet and/or other nearby devices.

A refrigerator, for example, may have a touch screen on the door and be connected to the Internet, allowing you to remotely access information — things like inventory, temperature, whether or not you have what you need to make a certain recipe. Another example is the Nest thermostat, which is a connected thermostat that allows you to remotely manage your thermostat from your smartphone, tablet or PC. The high-level view of the Internet of Things is a world where nearly every electronic device we own will be connected to something.

In years past, this idea was just an idea — something we said was coming. This year, however, was the first year when I could actually say the Internet of Things was on display. I saw examples of nearly every type of electronics device — from coffee makers, ovens, fridges, cars, clocks, stereos, exercise equipment, and my personal favorite: an LED lightbulb with a wireless speaker built in. All of these devices were connected to the Internet and allowed you to interact with them, store data, access data and more. This was the first year I could see the Internet of Things becoming reality, and it is very exciting for us industry observers.

I’m half-joking but I can’t wait for the year when we see a connected toilet and companion app.

Hardware with Software Accessories

The reality of the Internet of Things coming to fruition brings with it perhaps one of the most interesting developments: the role of software. What become increasingly evident with all the connected devices I saw and played with at CES was that nearly all of them were made significantly more usable and valuable through the use of companion apps for smartphones or tablets.

Some I have spoken with position devices such as connected watches or even the Nest thermostat as accessories to your smartphone. The logic is that because your smartphone is the terminal that all these devices leverage to really make them smart, then the phone must have the more important role. This is true to some degree, because there is no point of having all these connected devices if the experiences are limited to the hardware itself. However, I would still position the hardware as the stand-alone device adding value, while the app on your smartphone is actually the accessory.

At CES, this hardware-software model was on full display in the area of smart health and personal sensors. On the show floor, the pavilion for the smart/connected health area was the biggest I have ever seen it: Several dozen vendors were there showing off the latest in smart health. Every single smart health and body sensor product I got a demo of or gathered info on had a companion app that ran on a smartphone, making the hardware more than just hardware.

This is the world we are headed toward. Because of the unrivaled momentum and rapid worldwide adoption of devices like smartphones and tablets, we have smart devices with us at all times. They perfectly function as the platform to drive the interaction with the hardware around us.

All hardware will be made smarter through not just the use of connected chipsets and next-generation parts, but rather through the applications that add to their value.