At a very interesting ITEXPO (News - Alert) Austin 2012 panel discussion on October 4, 2012, the big topic at hand was big data. We all know about big data by now – yet another tech buzzword that refers for the most part not only to vast accumulations of structured and unstructured data that is now being generated at roughly five exabytes every two days (keeping in mind that as Google’s chairman Eric Schmidt (News - Alert) noted in 2003, the entire amount of data created since the dawn of our ability to generate data of any sort as cave paintings up to 2003 humans had created roughly five exabytes of total data to that point in time), but also of nearly endless streams of sequential “times series” data being generated by machines in machine to machine (M2M) ecosystems.
In addition to all of this data there is the other side of big data – taking it all and making sense of it. Converting big data into usable intelligence is a challenging task, and is really at the core of what “big data” is all about from a technology buzz word perspective.
Interestingly, the key message from the panelists at the Big Data session – which was specifically focused on M2M-generated data, was that a) there is much too much M2M data, b) much, if not most of it, never needs to be transmitted endlessly from the end points where it is gathered to the huge data pools in central repositories where it is stored, c) most of it never really needs to be stored for any particularly lengthy periods of time, unless, as one panelist said, “You really like paying the big data storage companies a lot of money.”
Rick Aguirre, president of Cirries, whose products allow communications companies to create a unified service layer across wireless, wireline, cable, video, broadband or IP networks, whether for Telecom, M2M or the cloud, presented an overview of big data. Aguirre pointed out that M2M data is invariably geographically dispersed and distributed, is generated in real time (suggesting that for the data to have the most value it needs to be evaluated as close to real time as possible – but only when necessary), and somehow needs to go through a mediation process in order to determine if the data is indeed necessary at any given point in time. Mediation in turn requires a level of policy application for making determinations.
Big Data is Much More Valuable as Little Data
That was all well and good and Cirries will deliver the necessary channels to get that big data delivered. Next up to take a whack (we do not use the word “whack” loosely here) at big data was John Caposa, general manager at Thingworx, a very cool company that recently emerged from sort of a stealth mode to launch a new M2M development platform at the end of 2011 that has caught some serious fire among the M2M developer community. Caposa pulled no punches – he is a serious believe that when a term reaches buzz word status things begin to get out of hand.
Caposa believes that there really is far too much M2M data in circulation, that it is crazy to worry about storing it all, that M2M data for the most part (leaving aside specialized applications where storing the data is of great importance – a small percentage overall of the total big data population) never really needs to be gathered unless there is a specific need to do so, and that if any company finds itself with a vast collection of big data it probably isn’t effectively making use of it quickly enough to turn it into highly valuable business intelligence/information. We couldn’t agree more, nor do we disagree one bit with the overall tenor of the big data discussion overall.
These points of view beg the question: How does a company take advantage of big data within M2M environments quickly and how can it be used to create real business value? Caposa left the question hanging in the air, but set the stage for the final speaker on the panel, Fred Yentz, president and CEO of ILS Technology, a company that specializes on M2M (as well as machine to enterprise, machine to human, and machine to cloud – we sense a trend here) connectivity between the edges of an M2M ecosystem and the BI capabilities in the home base (and data repository) of that ecosystem.
Yentz immediately echoed Caposa’s view that there is far too much “big data” being transmitted at tremendous cost (even in the case of utilizing 3G and even 2.5G networks) – for the most part for no reason what so ever, other than companies not realizing they do not need to. Yentz refers to it as managing bandwidth economics. He pointed out that in an age where being agile with data is critical, but that agility is being significantly reduced and in some cases smothered through the overall weight of big data. It’s a colorful image but it gets the exact point across.
Yentz noted that “The key to enabling big data as intelligence is to consume information only at the right time – right time data is what makes BI highly effective.” The ILS solution is to provide edge logic and filtering of all data. Keeping as much data at the end points, rather than transmitting it to a data store back home is a key means of reducing data traffic, eliminating the “big” in big data, and creating manageable data stores that can indeed be quickly acted upon.
One example Yentz provided is to store the sequential, time series data we noted above locally to the edges of the M2M ecosystem, and even then suggests that only storing a certain amount of that data is necessary. If an M2M device signals an alert, then a transaction can take place that initiates an actual data transmission from the edge to the BI machine for analysis. Depending on the circumstances that might result in a transmission of five minutes worth of data, or a day of data – whatever is appropriate for the specific application.
At this point the panel ran out of time, but the panelists certainly made their point exceedingly clear – keep that big data as far away as possible.
Want to learn more about the latest in communications and technology? Then be sure to attend ITEXPO Austin 2012, happening now in Austin, TX. Stay in touch with everything happening at ITEXPO. Follow us on Twitter.
Edited by Brooke Neuman