The Internet of Things and the Internet might seem inextricably linked, but, increasingly, there are questions centered around how IoT devices should work with one another — and what happens when the Internet connection goes down?
Users also are concerned with the privacy implications of having their data stored on a corporation’s servers, and they don’t like having an Internet connection as a potential point of failure. These reactions are rational, but reminiscent of online shopping circa 2000, which, ironically, might now be more secure than shopping in physical retail stores.
To understand why device makers are relying on an Internet connection and cloud services, we need to look at how our IoT devices work. We need to understand data sources, processing, device to device communication and, ultimately, how one device can leverage another device.
As a maker of climate control devices, there are only a few critical sources of data: humans, their environment (indoor and out) and energy utilities.
There are humans who have a desire to be comfortable, which boils down to having a certain air temperature, radiant temperature and humidity, among other things. Humans live in a variety of geographies, meaning there are often large differences between what they like inside and actual outdoor conditions. Imparting product showcase video comfort into a space with a large indoor/outdoor difference takes energy, and because energy is subject to supply and demand forces, using it intelligently means understanding its price at any given time.
Let’s distill these down to some concrete data sources. Phones, as arguably today’s ultimate wearable, are a source of data, including location, both macro level (at home or away) and micro level (in a particular room for more advanced systems like ours). They also provide information from human input, accelerometer movement and, in some cases, their microphone.
Does every flick of a light switch need to go to the cloud and come back down to turn on your light bulb?
Sensors provide environmental data (indoor and outdoor) about raw physical conditions and, often, because of cost and power constraints, have little-to-no processing onboard. In our case, temperature, humidity, ambient light and more all give a digital view of what is going on in a space. There are external sources of data, as well, like electricity and gas rates from utilities, or weather conditions observed by third parties.
Already, you can see cases where a system at home can operate most ideally with offsite data sources: phones reporting their location when you run out to get groceries, weather from Yahoo and utility rates from PG&E, for instance. This raw data leads to the next question of where it should be sent and how it should be used.
All this data needs a place to go. Although you could host a server farm in your home, you probably don’t want to invest in one — economies of scale dictate that Amazon will be much better at it, and configuring your router to accept data pushes alone is the stuff of consumer tech nightmares. An always-on, scalable cloud, redundantly backed up on Amazon, is much better than virtually anything you might practically have in your home.