Happy Iotday!

Happy Iotday 2018!

Today marks the recognition of one of the most promising technologies in the Industry 4.0 space - the Internet of Things. What makes it so important? The Internet of Things (IoT) or, in our specific industry, the Industrial Internet of Things (IIoT) is one of the base-level technologies which connects the physical world with the digital one. It seeks to connect physical data sources with the massive-scale computing resources available via the Internet in order to (hopefully) improve manufacturing processes and products which are the "things" in this equation.

Surprisingly, the basic concept of IoT is not really that new. For decades, manufacturing lines, production machinery and products have been decorated with sensors. These sensors would send data to a control system which could sometimes be networked. The control system would then monitor and perhaps adjust the physical process to make it better. Such adjustments were performed by "actuators" which is a blanket term for any device or human that can impact the physical world.

So, the "Things" have certainly been with us for quite some time.

What's all the fuss about then? You guessed it - the Internet! With the power of the Internet, the amount of sensors, the amount of data, the amount of inexpensive computing power and the geographical reach are all virtually endless. This has opened up new possibilities and products for commercial consumers and manufacturers.

Let's look at an example of a common IoT system - the mobile phone in your pocket with the Google Maps app installed.


Have you ever used Google Maps to look at the traffic in your city or for advise on how to navigate around heavy traffic? How does Google create the map above (downtown Chicago at 10:00 am CST)?

Among the myriad of sensors in a modern mobile phone, there is a GPS sensor. From periodically sending the GPS data (in the case of traffic velocity data) of the millions of people on the road to Google's infrastructure, Google can create the visualization above.

To capture millions of data points frequently, you would expect that the computing resources necessary would need to be massive in size. And you would be correct. But until recently, the emergence of Cloud Computing has given companies, small businesses, startups and individuals the power to utilize massive amounts of computing resources relatively cheaply. In the past, if you were a small business or a startup with limited capital, you would have to acquire and maintain large, on-premises servers to accomplish the same task - a deal breaker oftentimes.

Along with your GPS data, Google also uses the internal Cloud Computing infrastructure to look at past data to develop predictive traffic models based on the day of the month and the time of the day in a certain area. Cloud Computing comes to the rescue here as well! For machine learning (a specific component of Artificial Intelligence), large amounts of computing power are needed to ingest data, develop and train models and perform inference.


And we are not done yet! It turns out that Google had also acquired a company called Waze a few years ago. Waze is a mobile app which allows users to report road conditions while they are traveling (hopefully with a passenger who can safely operate the app). Users can report potholes, construction areas, accidents and the presence of law enforcement.

Believe it or not, Google also uses this user-supplied information to impact the traffic visualization in Google Maps.

Did you notice something interesting? What is the "sensor" in Waze? You are! Waze relies on humans to "sense" the road conditions and report said road conditions via the Internet.

We call all of the above data fusion. This involves the combination of a variety of data sources to tell a story or to paint a complete picture of some aspect of the physical world. Let's review the data sources involved in this data fusion:

1. Live GPS sensor data.
2. Predictive data from past GPS sensor data.
3. Live and past user-reported road condition data.

Can you think on how this same architecture can be applied to the typical factory floor?

1. Live GPS sensor data. Machine tool status information via MTConnect or OPC-UA.
2. Predictive data from past GPS sensor data. Predictive data from past machine tool data.
3. Live and past user-reported road condition data. Human supervisors or floor personnel reporting factory floor conditions via HMIs or floor workstations.

In the same way Google builds a visualization of traffic conditions, we too can build a visualization of factories - and increasing complex visualizations at that.

If you want to learn more about IoT and IIoT, I would recommend my recent talk on the SME Virtual Network - IIoT Fundamentals and Applications. Consider subscribing to the SME Virtual Network YouTube channel as we will be discussing this in more detail in the coming months. If you have any specific questions, feel free to message me or reply right here on SME Connect!

Have a happy Iotday!


Adam J. Cook
Chair of SME Chapter 112
Posted by Adam Cook on Apr 9, 2018 12:02 PM America/New_York