In this article about the Internet of Things, published by MSDN you’ll learn how to leverage the Windows Azure Service Bus, not only to collect information from “things” but also to control them.
When you stroll (or browse) through a well-stocked electronics store these days, you’ll find an amazing array of “things” that have the ability to connect to a network. Don’t just think phones, tablets, notebooks or desktops, or just TVs, Blu-ray players and set-top boxes. Think espresso makers, refrigerators and picture frames. Think garage door openers, air conditioning and alarm systems. If you look around and behind the cover panels in industrial or commercial environments such as your own office building, you’ll find temperature and humidity sensors, motion sensors, surveillance cameras and a multitude of other kinds of sensors or control switches inside equipment.
Many of these devices generate useful data: temperature readings; number of cups of brewed coffee and how the grinder has been set; infrared images showing that no one is in the conference room and therefore the lights can be turned off.
It’s easy to imagine a scenario where you’d like to upload some data into a “thing” as well, such as pushing the latest pictures of your children (or your pet) to a picture frame sitting on grandma’s sideboard; or one where you want to flip a switch from a distance—even if that distance only means your phone connected via 3G/4G mobile carrier network—to turn the temperature in the house up a notch. From a networking perspective that’s three worlds away, but from the consumer perspective there’s no appreciable difference between flipping the switch at home or while sitting in a cab on the way back from the airport returning from a two-week vacation.
The opportunities around connected devices are enormous. Supplying services for special-purpose devices might indeed provide more monetization potential for forward-looking cloud developers than apps on general-purpose screen devices tailored for human interaction, such as phones, tablets or the many PC form factors. This seems especially true when you combine such services with cloud technologies emerging around “big data” analysis.
The Scenario
For the purpose of the following architecture discussion, let’s imagine the offering is a Software as a Service (SaaS) for air conditioners. While the scenario is fictitious and so are all the numbers, the patterns and magnitudes are fairly close to actual scenarios that the Windows Azure team is discussing with partners and customers.
What’s nice about air conditioners—from a business perspective—is that there’s healthy demand, and global climate trends indicate they won’t be going out of fashion anytime soon. Less nice is that they’re enormously hungry for electricity and can overload the electrical grid in hotter regions, resulting in rolling brownouts.
The SaaS solution, for which I’ll outline the architecture, targets electricity companies looking for analytic insight into air conditioner use for the purpose of capacity management and for a mechanism that allows them to make broad emergency adjustments to air conditioning systems hanging on their electricity grid when the grid is at the verge of collapse.
The bet is that utility customers would prefer their room temperatures forcibly adjusted upward to a cozy 80° F/27° C, rather than having the power grid cut out, leaving them with no defense against the scorching 100° F/38° C outside temperatures.
Let’s further assume the SaaS will be paired with a number of air conditioner manufacturers to integrate the required hardware and protocols. Once the devices are installed and connected to the service, there will be opportunities to sell electricity-company- or manufacturer-branded companion apps through mobile app stores that allow the customers to monitor and control their air conditioners from their phones or tablets.Next figure shows an overview of the scenario.
.png)
