Domino’s Rule is simple: if you can’t get a pizza delivered, you probably can’t count on broadband.
It is a very useful rule of thumb. For example, in the Permian Basin’s Wolfcamp formation, roughly two miles below the surface of West Texas, you will not be able to deliver a pizza.
However, two miles down, what you will find are ESPs; electrical submersible pumps. At that depth, the pump and installation will cost well over a million dollars, and can easily exceed two million. And, at that depth, you will not have a broadband connection to send data out of the well. Even if you could get broadband data up the well, the chances are good you won’t have broadband fiber or wireless connections at the wellhead. You can’t get a pizza up there, either.
Another example is the airport jet bridge. If you have ever tried to use a broadband connection on an airplane before takeoff, you know it can be hit or miss. The cell tower is far away. Suddenly, you are transported back a decade or more. The data stream is S.L.O.W.
You are not the only one having a problem. Modern jetliners land with 100 Gigabytes to half a Terabyte of data. But with just minutes to turn the airplane around, most operators simply dump this data. You can’t get a pizza delivered to the jet bridge. You also can’t count on broadband to ship the data out.
A third example is rail cars. They roam over vast expanses. You can’t order a pizza delivered to a siding in North Dakota, or Siberia. And, an IoT device on your rail car can’t reliably ‘phone home’ there either.
All of this is frustrating today, but it seems likely to be getting worse. IDC estimates the global “Datasphere” is roughly 20 Zettabytes in 2017. That number will grow by a factor of 8 to 10 by 2025. The IDC study can be found here; www.seagate.com/www-content/our-story/trends/files/Seagate-WP-DataAge2025-March-2017.pdf
A great deal of that data will be generated at locations impacted by Domino’s rule. Even if you can get data (or pizza) delivered, do you need to pay for all that data transport? It can be expensive to transport data at an industrial scale.
What does that mean?
It means processing must move to the edge. Most datasets are created at the edge (or at the “endpoint” depending on your semantics). Unless we want to keep dumping most of the datasphere, analytics must move to the edge.
So, for edge analytics success we suggest three things:
- Try to order a pizza
- If you CAN get one, we’d suggest either a tech enabled solution we can provide using our TruNavigator analytic environment; or, an application hosted in our AnaltyicsOS Server Edition may be what you need.
- If you CAN’T get a pizza, you need an application with true edge analytics and THAT is AnaltyicsOS.
P.S. We are grateful to Domino’s for the use of their photo, and their good-natured response to “Domino’s Rule.”