In a recent conversation with Forbes, Rishi Chandra, Vice President of Product and General Manager at Google Nest, defined a subtle but important change in terminology.
Speaking to Forbes Chandri said: “We’re intentionally deviating from the word smart home, because we actually think it’s the wrong word, we actually do think it’s a very tech-oriented way of thinking about the home. No one asked for smartness, for the smart home. That’s not an attribute necessarily people care about. What they want is help. Like, we want to go from what a technology does to where it actually provides benefit. And so our mantra for the next five to 10 years is going to be the notion of how we can help deliver the helpful home.”
In the past I’ve written about the need for integrators to deliver smart home systems that provide their customers with, what I referred to as, “true automation” vs. a system that just provides another way for customers to control the lighting, HVAC, TV, etc. in their home. Smart speakers and smart phones simply provided another way of turning on a light switch without having to walk across a room. While this was “cool,” it really didn’t provide a great deal of value to most homeowners.
Chandra’s comments show that manufacturers are looking to take the next step and move beyond simple control systems. My expectation is that this change will leverage the power of machine learning and drive the voice recognition capabilities of smart speakers to move beyond the types of automation that integrators have been able to provide today. For example, an integrator might have programmed a customer’s system to:
- Turn off lights, TVs, music, and set back the thermostat to save energy when the home isn’t occupied
- Restore the thermostat to the normal temperature when people come home
- Turn on pathway lights when people arrive home at night
- Turn off the water to the house and sound an alarm when a leak is detected
- Disable the irrigation system if it rains
- Turn on an air cleaner if the air quality in a home is poor
- Remind the homeowner when the clothes dryer finishes drying a load of laundry so they can be removed from the dryer and won’t wrinkle
- And the list goes on…
A “helpful home” based on machine learning can provide the same functionality, and more. For example:
- GPS data from the car, or a smart phone, could show that the family had been to the grocery store so instead of a fixed set of pathway lights being turned on when they arrive home; the lights to the kitchen would be lit.
- The air quality and temperature in a home could be driven by sensors to provide clean air at a temperature that provides the highest degree of comfort. For example, the home might detect that exercise equipment in a home gym is being used and lower the temperature to keep the person exercising cool. On the other hand, if someone is lying on a sofa reading a book, the temperature might need to be warmer for the person to be comfortable. Cooking releases VOC’s into the air so an air cleaner could be turned on when the stove is in use.
- The wake up time of the homeowner’s alarm clock could automatically adjust based on events on their calendar and deadlines defined for items on their “to do” list.
- Instead of the fixed set back schedule that thermostats offer today, the thermostat could:
- Bring the house up to the normal temperature just before the homeowner morning alarm was scheduled to go off
- Set back the temperature to the night time setting when people actually went to bed
- Recognize when guests were over and make adjustments so everyone was comfortable. For example, during a summer party people frequently move between inside the house and the backyard. The system could sense that the door to the backyard was opening and closing, letting in air with lots of pollen, and engage the air cleaner.
While this utopian dream would have sounded like science fiction only a few years ago, I expect we will be seeing many of the above features sooner than we think. One challenge will be creating a system that can initially be helpful to homeowners and grow even more helpful over time. Using one of the above examples, the “helpful home” should, after installation of a machine learning system, turn on a default set of pathway lights when the homeowners arrive at night. Over time the home can then learn more about the homeowners movement habits and make better decisions as to which lights should be lit. How a machine learning system is provided with default behavior based on a customer’s unique home will be an interesting challenge.
It should also be noted that this utopian vision also has downsides
Privacy
For a “helpful home” to operate it will require a large number of sensors that monitor the activities of the people in a home. How this data will be used raises significant privacy issues. The E.U. is far ahead of the U.S. in laying a foundation of laws to govern personal privacy. The data collected by a “helpful home” is going to raise the stakes in how data collected, about the way we live our lives, in the privacy of our homes, is used.
Security
In the “helpful homes” more and more devices in our homes are going to be interconnected including:
- Light Switches
- Thermostats
- Appliances
- Thermostat
- Air Quality Monitors
- Motion Sensors
- Cameras
- Health Sensors
- Irrigation Controllers
- Water Heaters
- Audio and Video Equipment (TV’s, etc.)
- Security System
- Door locks
- And More…
Protecting all of this from hackers is going to be a significant challenge. It would be very annoying for a bored teenager down the street to have fun turning your bedroom lights on and off in the middle of the night. But, it would be a much more significant issue for your thermostat to shut off in the middle of the winter and to have to pay a ransom to regain control of it before your pipes froze/burst, causing thousands of dollars of water damage in your home. Just as machine learning will help make our homes more “helpful” this same technology needs to be applied to our home networks to detect intrusions and make them more secure.
Cost
How a system that creates a truly “helpful home” is paid for hasn’t been defined. Starting with smart speakers, the systems in our homes are relying more and more on powerful cloud servers to run them. The sale of the data collected about us has kept people from having to pay for the cloud services we use today. Whether this continues in the future, or whether laws that provide privacy are passed, requiring companies to pass the costs to consumers in the form of subscription fees, remains to be seen.
Summary
Rishi Chandra’s coining of the term “helpful home” is a step in the right direction towards home systems that are focused on customer needs instead of tech innovation. However, privacy and security safeguards are going to become even more important than they are today for the “helpful home” to be successful.
