Every other article on IoT will have a stat reporting on the size of data being generated by IoT—currently estimated at 400ZB (ZettaBytes) in 2018 (Cisco Cloud Index Study). A Boeing 787 airliner generates half a terabyte (500GB) on every flight. A zettabyte is a trillion gigabytes, and as such is almost meaningless, or certainly impossible to visualise—much like a nano second or even a microsecond. People seem to love using numbers almost for numbers sake.
Admiral Grace Hopper (1906-1992) of the US Navy infamously called up her engineering department and asked them to “please cut off a nanosecond and send it over to me”. What she was asking for was a foot of wire—the distance through which electricity travels in a nanosecond. She often said she would like to hang a thousand ft. long wire over every programmer’s desk—or around their neck—so they knew what they were throwing away when they threw away a microsecond of time (60 Minutes interview 24 August 1986).
It’s important to get context in numbers—to make something big (or small) meaningful, just like Hopper did with a nanosecond of wire. Just because something is big, doesn’t necessarily make it good. It is what you do with it that really counts.
There are two traps waiting to snare the foolhardy rushing headlong down this new and exciting road of using IoT and the data it generates. The first being the staggering rate at which the data is being created—more data has been created this decade than in all those preceding. The second one is that of doing only something mundane with the data and believing it is actually useful in its own right.
Apart from avoiding the traps, the critical challenge is using the data when it is still in flight and going beyond just extracting the supposedly valuable information.
Data is not gold, rather it is the new oil, because like oil, it is worthless until it has been refined.
The real value comes from doing more than just ‘historical analysis’, which was the task of the first generation data miners who often used the data to simply determine ‘what happened’. The size and sheer volume of today’s data might be seen as a hurdle hindering the refinement process. However the enlightened miners are evolving, and soon we will get a new generation whose mission is more than display and respond. They are not daunted by bulk and size, they can get the small, micro nano detail and are driven to do something that counts.
A change of focus is required, to move to a mission of using data in a way that really counts—to “predict and act”. The second-generation miners with some generic predictive tools are now able to start giving answers to “why did it happen?”. Our third generation miners will bring a higher level of focus bringing true predictive modelling which will allow us to get answers to the “what will happen?” question. Fourth generation miners will use data for simulation and optimisation and so will be able to answer, “what is the best that could happen?” deriving real efficiency and maximum benefit from the plain ordinary data.
We should therefore be basing our IoT strategy around an ecosystem of data, devices, solutions with actionable analytics and looking forward to the generation that have a mission to ‘predict and act’, moving right away from the ‘what happened’ scenario.
The England rugby team have done just that in their record equalling 18 winning streak. A GPS unit in their jerseys provides the data measuring the number of seconds it takes each player to rise from the ground. The players have a target of a of maximum 3 seconds. Eddie Jones says England are getting better but are still ‘seven per cent slower’ than the All Blacks, adding that in the bad old days, ‘some of the blokes had a cup of tea and a scone with jam and cream before they got off the ground!’.
Actionable analytics used in the right way (and IoT in the background) is helping England train more efficiently and as a result play more efficiently.