The Internet of Things is coming. If you don’t know what that means, there’s an explanation here. But if you want to be brought up to speed more quickly, here it goes:
We’ve gone through three waves of data creation in our modern age. In the first age, data was generated by back office systems as we first computerized. Enterprise applications gradually replaced those first humble machines and information technology was off and running. The second wave came from the arrival of the Internet, which brought us whole new human sources, including social media data, publishing at will and new IT challenges around volume, velocity and variety.
But it’s the third wave that is the real tsunami. The third wave is machine generated data.
Things creating data
The tsunami that’s gathering force is powered by data created by ‘things’…objects and machines like servers, gateways, and sensors. These objects are getting very small, very smart, and very sensitive to things like pressure, light, temperate and movement. They’re also becoming very configurable and can take instructions through wireless signals that allow for changes in rules, data handling and events. Lastly and most importantly, these objects are always on and always connected.
According to IDC, “In 2012, ‘machine-generated’ data represented 30 percent of all data created, up from 24 percent last year and 16 percent five years ago. With machine/device data volume increasing rapidly, the number of connection points is also moving from billions to tens of billions. IDC predicts that by 2020 machine data will rise to 40% and will have more devices than people on the Internet.
Once our parking meters, cars, wearables, and even our refrigerators are data generating and consumption ‘things’, new approaches will be needed. The systems we’ve built for the first and second waves won’t react quickly enough and in that decision latency comes risk, dissatisfaction and lots of lost business value.
Richard Hackathorn developed an excellent graphic that shows the lost value of decision latency, borrowed here from SmartData Collective.
The “Time is Money” curve is a challenge as we’re becoming more reliant on systems and networks to provide instantaneous feedback and often, gratification. We clearly need to continually work to shorten data gathering and analysis, but the response time, either through automated or human-involved process, is the heavy lifting.
This means the three biggest challenges for Big Data are extreme scalability, real-time event handling, and time-to-insight. To be successful with any Big Data project, all three need to be central drivers.