I have a good friend who hates the term Big Data. She always says its not the size of the data but the analytics that matter. I agree with her. Large data sets are more driven by the ability to analyze the information than by the actual size of the data. As we head down the CPS path there are additional things to think about in terms of data. Data has a time to live. I made my joke a couple of days ago about the open manhole cover warning system. Really good information to have as you walk along the sidewalk on your way to work. Not information that does you any good as you are falling into an open manhole cover on your way to work.
The when you get data is critical. Two seconds after you hit the bottom of the manhole with a broken leg the data is useless. (its sexist by the way to assume only men should go into sewers, lets change the name to personhole). Data that is too large for your device to consume is also not relevant. A map with all the open personhole covers in a city on your phone is useless if you are 1 foot in front of a currently open personhole.
The legendary interview question why are all manhole covers round. The answer – they aren’t. Data is time driven it is also device driven. Both the producing device (CPS or other) and the receiving device (laptop, mainframe, desktop, cellular phone, server, bank of servers, cloud provider). The reality of the data as much when it needs to be consumed as well as where it needs to be consumed.
CPS devices today (2015) produce around 110 zB of data. Where z represents Zeta. A zeta byte represents 1000 Petabytes of data (pB). It’s a lot of information. Some of it can be quickly discarded (you only need the video feed relevant to the problem, the rest of the feed can be discarded). But that is today, October 2015.
A little over two years ago a number of interesting campaigns started on Kickstarter and Indiegogo. This image is of the Bubl camera. 3d camera in fact. It very quickly lets you render a 3d image of any space you place the camera. It’s a lot more data than you produce with even a high end DSLR. It also adds an interesting twist to the data you collect, parse and use. If you get a chance to get one, I highly recommend it.
The data collected is changing. The presentation mode of the data is changing as well. Where you consume and how you consume data is radically different in just the last 10 years. The birth of BYOD changes how data is presented. The flexibility to work remotely or at the local coffee shop makes anywhere a work place (don’t get me started on the overall security threat that represents). Data now adds a third dimension that device specific actionable. An excel spreadsheet sent to me while I am wandering around with my iPhone as my primary connection isn’t a good use of my time, your time or the data and frnakly my device. A smart device system would notify the sender that in fact the user was on their cell phone and would not be able to interact with the information.
My friend who has long been ahead of the data importance curve was right. Its not big data. Big implies size. Its smart device data and analytics.