Over the past couple years, I’ve seen the proliferation of some new terms. The ideas don’t seem new to me – hell, we’ve been doing the tasks these definitions describe for a decade or more and have a full product suite based on the ideas – but evidently some corporate management teams seem to have gotten interested, and so they needed catchy names, with alphabet soup acronyms. Here’s my take on what some of them mean.
Machine to Machine (M2M)
M2M seems to be all the new buzz. Oracle thinks it will be huge (a $25B market by 2015). Intel thinks it will be even bigger (15 billion connected devices by 2015). But really, the concept is pretty simple, and it’s really not new at all. M2M is simply two machines talking together. Typically it’s low-powered (resource wise) things. So it’s not your PC talking to a server, but it might be a sensor talking to an embedded device. Or two embedded devices sharing data. What’s new is that we’re finally at the cusp of technology and cost where we can actually start fielding widespread M2M capabilities.
Moore’s Law is has finally gotten us to where pretty small devices, using small amounts of energy and costing very little can do a lot of things. Not so long ago a municipal bus couldn’t tell you anything about itself. Then it could tell you about fault codes if you brought it back to the bus barn. Then it could tell you where it was. But now, you can put a small appliance right on the vehicle that can tell the driver if a problem requires returning to maintenance. It can even call maintenance and give them diagnostic info or even tell them what part to have ready. Basically any device out at the “edge” of an overall solution that is capably of doing more than just collecting sensor data is an Intelligent Device. Obviously there’s a pretty wide range of what can be done out there, depending on your solution’s ability to absorb size, power and cost.
Maybe the device does nothing but monitor a temperature and then send a notification when it exceeds a specific set point. That’s intelligence when compared to a device that just sends the temperature to a server where the comparison and notification is generated off-device. It means that the device is capable of using Analytics to make a decision like when the temperature exceeds 205F, send an alert. It’s capable of using Aggregation to summarize, filter and/or generate new data like the alert condition itself which the temperature probe doesn’t have. It may be able to run Rules such as when the temperature is over 200F, start storing the oil pressure and passenger count. It may be able to Publish any of that data to another device or to a server somewhere.
It’s this intelligence, and really the ability to distribute different parts of this intelligence across a many devices and machines in a solution that has started changing the landscape of what can be done. Before we had a machine shop where operators would report part counts at the end of a shift, and a manager would create a spreadsheet that they would send off to the customer at the end of the week or month. Now, the machines themselves can report exact part counts in real time right to the end customer. That is enabling a lot of cool things.
The Internet of Things (IoT)
The Internet of Things, or IoT, in my mind anyway, is the collection of all of these Intelligent Devices and how they use M2M to actually send useful information (as opposed to raw data) between one another.
Let’s look at a concrete example.
Let’s say I have some sensors in an apartment building that give current temperatures. These report back to an Intelligent Device. This communication may be M2M (or it might just be a thermocouple wire – let’s not get too hung up on it).
That Intelligent Device can record that data. It can run rules on that data to determine if it needs to turn on or off the boiler. It can send a signal to another device to actually power up the boiler – again this might be M2M. The Intelligent Device can do aggregation of data, providing hourly rollups of data and even calculating heating and cooling curves for the building it’s in. It can also publish that data to another device or to the cloud (we’ll talk more about this magical “cloud” word in another post). Yet again, here’s more M2M.
Now let’s say the Property Manager has 10 buildings. They can connect to the cloud and look at a history of the temperatures in an apartment. They can send commands from their office to the Intelligent device telling it to change a set point, or just override the current rules on the device and turn on the boiler. All of this interaction is an “Internet of Things.”