This blog is by Claudio Frascoli from the Networks business of Nokia.
Predicting the future sounds like magic, doesn’t it? How then, can we associate magic with the concept of big data?
Before pulling a rabbit out of a hat, let me state for the record that big data really is just that: a lot of zeroes and ones with very little magical appeal. I truly believe the real magic doesn’t lie in those numbers as much as it does in the way we interpret them. Additionally, the path to the effective use of big data analytics is long and often complicated. Not too much magic there, either, and a good reminder that magic is often just the consequence of a lot of hard work, the right mix of skills, and a clear vision.
However, as big data becomes mainstream in telecom, more and more operators have the right level of skills, a data-driven organization, and powerful tools in place. Therefore, we start to see the next wave of innovation coming to the fore, with use cases that exploit the power of data analytics to give us a glimpse of the future.
I was talking to a South East Asian operator this week and we were looking at use cases where big data can help improve its operations. “I want to be able to move from scheduled to predictive maintenance. I want trucks to roll only when I know my telecommunications equipment really needs attention.” This is what logistic companies like UPS have been doing for years with their fleets, equipping each truck with tens of sensors and using collected data to predict life expectancy of each component and adjust maintenance schedules accordingly. Applying the same principles to telecom equipment will require a shift in the way products are designed, but there is no reason to doubt that this will be reality in the not too distant future.
More broadly, big data analytics allows us to make sense of a multitude of data that historically has been discarded by telecom operators. When the right skills and tools are in place, we can use past data to discover patterns and link them to network and customer behaviors in ways that only a few years ago were hardly imaginable*. And better still, we can also add new dimensions by correlating information that was not previously considered. How does weather impact people’s movements and behaviors? How do those affect mobile traffic and, ultimately, the performance of a network? Understanding past patterns is indeed the first step we need to take if we want to answer to these kinds of questions and effectively predict the future. Or, at least, predict how our future service experience could be.
Of course, the ability to predict is only useful when we have the ability to act on the information in a way that will modify the final outcome for the better. Detecting events before they potentially affect service is good; being able to put corrective measures in place before the event actually happens is better. Determining how the current performance of the network can lead to customer dissatisfaction is good. Proactively changing course so that customers never get to the point of dissatisfaction is better.
Predictive analytics is an extremely powerful tool and what we are seeing today is just the beginning of a trend that has tremendous potential. For now, this is as close as it gets to time traveling. And it looks pretty magical to me.
Please share your thoughts on this topic by replying below – and join the Twitter discussion with @NokiaNetworks using #BigData NetworksPerform #mobilebroadband #CSPCX #MBBFuture #Nokia.