Caregivers and the “Smart” Homes of Tomorrow

This article features comments I posted on a James Holloway article about Smart Homes of Tomorrow, where automation is based on sensors and learned intelligence that encompasses any device providing automatic control of home functions. Systems most likely to be automated are: lights, thermostats & home appliances; television, video & music systems; security alarms &  monitoring systems; and home health care monitors, alarms & communication devices.

A conceptual smart home with 17 components, including automated pet feeder.

mHealthTalk Comment:

My perspectives aren’t too far from what Mr. Holloway wrote about. They came from introducing IBM to the Smart Home market in 1994, helping it launch IBM Home Director, and retiring in 1999 to start CAZITech, a Digital Home consulting firm.

The “Smart Home” concept has been around for some 40 years, but it never was able to cross the chasm from an industry niche of professional installation in high-end custom homes to mass market consumers. One reason is that any new device or change of behavior required a technician to come out and reprogram the house rules again.

As Holloway describes, a truly Smart home is one that’s aware of its surroundings from various sensors for light, temperature, sound, smells (smoke or CO2), motion, etc. A Smart home learns on its own about the various smart devices, occupant behavior and preferences and adapts automatically or upon command (by voice, button or gesture interface). And a Smart home communicates in ways occupants prefer, on a device they have at the time. But there are many problems with even that vision. A big one is that automation costs and effort aren’t justified in many cases.

When I moved to Austin, I built a 3,200 sq.ft. two-story home as a living laboratory to study the technologies and my technophobe wife as typical of consumer reactions. The local paper did a 2-page feature article on my home, which at the time was automated with X-10 ActiveHome, the precursor to IBM Home Director. Several (rules-based) subsystems were integrated and automated, including security, HVAC, lighting, fish pump, etc.

Later, after retiring and growing empty-nested, we downsized to a much smaller one-story home, and even though I installed the structured wiring, I never implemented much automation. At bedtime I can now just glance around and see if the doors are locked and lights on. I no longer need that one button to put the house to sleep, set back the thermostat, arm the security, and turn off the lights. It just isn’t justified in the smaller home.

The new NEST thermostat comes close to learning consumer behavior, but it doesn’t know if I’m cold because I just finished eating ice cream or if I’m hot because I just finished vacuuming. Without sensing that through some sort of skin patch to measure surface temperature, no thermostat would ever get it right. I adjust mine manually several times a day since it’s just steps away. Like many consumers would say, “I don’t need no automation.”

Besides the fact that automated homes have too many devices and sensors and networks from too many different manufacturers with different network protocols business objectives, there’s no clean way to show consumers the value proposition of an integrated Smart Home. Retailers like BestBuy learned to combine big television screens, electronics, and surround-sound speakers so users could “experience” the benefits of a home theater, but that’s nearly impossible to do with Smart Home technology, especially when artificial intelligence learning is applied. Although I know how to get around that problem, I’ll stop now rather than make this comment longer. ‘Hope  this perspective helps.

To learn more about home automation and my own home, read Smarter Homes for Home Health Care.


5 thoughts on “Caregivers and the “Smart” Homes of Tomorrow

  1. OK, I’ve got to add a comment that I also posted on Halloway’s article when someone was having a rant about how great the NEST thermostat was.

    My comments were meant to support the author’s premise for his article, which ponders how smart a smart home should be before it’s worthy of the name. Admittedly, NEST is smarter than a manual thermostat and smarter than most programmable thermostats. They all can sense temperature, but what makes NEST unique is its cool design, a proximity detector, and the ability to learn some patterns of consumer behavior. But NEST doesn’t integrate with other home subsystems (yet), so it can’t sense other things. It can’t sense a fire and shut off the HVAC to avoid spreading poisonous gasses around and fueling the fire. Even the nonprogrammable thermostat in my old home could do that, because HVAC, lighting, and security were all integrated, although programmed with a set of rules.

    My automated home, with its integrated but old-style thermostat, also contributed to significantly lower monthly energy bills that were half of what they were in my previous house of half the size. But other factors contributed too. The new home was newer, with better insulation and more efficient appliances. And the city-owned utility in Austin charged less per KWh than the private utility in Dallas. My point is it’s often hard to make good apples & oranges comparisons.

    I also stressed that smaller homes need less automation and “smarts” than larger ones. Through my work at IBM, I was involved in projects that were much further along the Intelligent Agent “Smarts” curve than NEST. My favorite never made it to market but would have blown the doors off of anything I’ve seen since, over the last 10+ years. Why it never made it is another story.

    Where should sensors reside, and where should learned intelligence reside? Sensors can be put into everything with a digital heartbeat and even in dumb devices with no power, by using RFID and low-power wireless like ZigBee, Z-Wave and Bluetooth-LE that can power sensors for years on a single, hearing aid-sized battery. But should the intelligence reside there too? Should the thermostat keep its learned behaviors to itself or share?

    From an engineering point of view, intelligence can go anywhere, but for maximum benefit, all devices should have access to it. Now imagine a Smart Home that has dozens or hundreds of sensors, has learned occupant patterns, personal traits, and ways they like to communicate. And now imagine that years worth of that recorded knowledge is stored remotely in a trusted Internet service that the house, or an authorized user, can tap. The company that implements that vision can established a huge barrier to entry since competing products wouldn’t have the benefit of “experience” with the occupants. The occupants also wouldn’t be tied to an automated home with the intelligence stored locally, because it could easily follow them to the new home.

    With so many big players in this space (Cisco, GE, IBM, Intel, Microsoft, Panasonic, Philips and more), one would think that one of them would have figured out how to take this niche market concept mainstream by now, after 40 years. But they haven’t. That’s because there are many, many missing pieces in the Smart Home vision, and most of them I never mentioned. It’s just not that easy, but we’re getting there, one smart device and network connection at a time.

  2. Dear Wayne, Werner and others,
    SMART is out since it has been used for too many dumb things… (plastic Xmas tree is now smart because it blinks when you get close etc etc)
    What i’m missing in the previous remarks is the link with the ‘living ones’
    How can you expect assistance from a system if that system has no clue of who you are and what you’re doing.
    This context-related approach is what we are testing now for a future development in AAL.
    Jacob D’Hollander
    Designer of assistive technology.

  3. In a press release this week, ABI Research said, “90 Million Homes Worldwide Will Employ Home Automation Systems by 2017.” I think that number is grossly inflated and responded back to tell the analyst why. Here’s what I said:

    Re. “ABI Research’s ‘Home Automation, Security, and Monitoring’ report examines the evolving marketplace for home automation systems and how the rise of managed subscriptions will impact existing luxury, DIY, and mainstream home automation markets.”

    While I’ve not read the report, I can agree that subscription services will drive more home automation in luxury and DIY markets. But I see no way HA will have much impact on mainstream homes. And I see no way to extend home automation to 90 million homes by 2017 without narrowly defining managed “home automation” as the use of any technology to automate one or more subsystems, such as by installing a programmable thermostat or remotely monitoring a security system. Such a narrow definition, with no mention of integrating different subsystems, misleads the public and the industry.

    To sell their reports, market researchers have inflated forecasts in the past by narrowly defining markets. When estimating the number of home networks, for example, they’d often include homes with just a broadband router and nothing other than one PC plugged in.

    Home Automation is more than a narrowly defined application of technology, and the concept has been around for some 40 years without crossing the chasm from an industry niche to mass market consumers. There are good reasons for that.

    I referred back to this article for more perspective on  that opinion, which comes from many years worth of both writing and purchasing market research reports.

  4. I don’t buy into the Smart Home hype started by ABI Research and picked up by so many other “analysts,” and here’s how I responded to Hurdles slow wireless home automation, mHealth, a recent article in Fierce Broadband Wireless.

    Home Automation and “smart” home systems are less justified in smaller homes, as I learned when my empty-nested wife and I downsized to a smaller one-story home. Because I had once introduced IBM to the Connected Home concept, used my own home as a living laboratory, and helped develop IBM’s Home Director (long since spun off as a separate company), I made sure the new home had all of the structured wiring features to cover whatever I may want to add later. But since moving in 5 years ago, I’ve added nothing, nada, zilch.

    Where I once pressed a single button on the bedside night stand to turn off all of the lights, set back the thermostat, and arm the security system, I no longer needed that convenience and found it more of a headache. When retiring for bed, I now glance here then there to verify that the doors are locked and lights off. (I don’t have to run up & down stairs.) I even decided “not” to install the NEST learning thermostat since I know better if I’m hot or cold.

    I do think you should make a distinction between smart home technologies and smart homes and settle on a standard definition of “smart”. Home Automation, which is accomplished through programmed rules that must change whenever a new device is installed or living habits change, can hardly be called “smart.” The NEST thermostat gets close to the definition since it has some ability to learn, but it doesn’t interface with the lights and security system to turn off during a fire, for example. A smart home would respond to a vast collection of sensors to give it an awareness of what’s going on inside and have the ability to learn and adjust to its occupants automatically and without programming.

    Apparently I’m not the only one who’s skeptical. Pew Research Center just reported that “Experts think tech-enhanced homes appliances and utilities will spread,” but more slowly than suggested by marketing hype. See Will Healthcare Lead The Future of Smart Homes?

Comments are closed.