July 22, 2016
By 2021, out of 16 billion IoT devices, 1.5 billion will have a cellular subscription and will be generating an infinite amount of complex data that will require a different approach for efficient analysis and use.
As Cheryl Wiebe, Partner at Teradata, explained, As products become increasingly sophisticated as computing devices in their own right, so does the data they produce—the connected car, the connected wind turbine, the connected train, the connected oil rig—each producing accelerating amounts of data.
Operations of Things and the Analytics of Things (AoT) are distinguished by Wiebe as the two pillars of IoT aimed to maximize the value of the world of connected devices. According to Wiebe, The Analytics of Things subsystem of IoT is an extension of the company’s analytic ecosystem that includes:
The complexity and variety of data increased over time with advanced sensory technology adopted across industries (wearables, cars, buildings, mobile devices, etc.). The tools traditionally used by businesses to analyze the data flow are no more valid as the stream widens and diversifies. According to some estimations, 2.5 quintillion bytes of data are created every day with 90% of it being unstructured.
For businesses to be able to make effective business decisions, the ability to manage the stream of meaningful data is crucial. Intelligent management of sources of data from various divisions of organization, its assets and partners can enable managers to make efficient data-driven decisions and see organizations succeed in the competitive environment.
Tom Davenport, the President’s Distinguished Professor of Information Technology and Management at Babson College, distinguishes predictive analytics as one of the four (along with descriptive, diagnostic and prescriptive) typology of analytics of things.
Among the examples of predictive analytics for the IoT, Davenport brings up predictive locational analysis, which happens every time the smartphone or car GPS are used to plan a route. Another example lists increasingly common predictive maintenance on industrial machines, which alerts companies that their equipment is about to break down so that it receives service in a timely manner.
Predictive analytics as a typology of AoT can also have a significant impact on the healthcare industry. As Davenport explained, Predictive health is another area with a lot of potential, but not much actual, value. Applications could take your daily steps, weight, and calorie consumption (that’s the toughest data point at the moment, since it relies on self-reporting), and predict things like your likelihood of getting Type 2 diabetes, or even your lifespan. More prosaic predictions could involve your likelihood of losing weight in time for your class reunion, or you beating your best time in an upcoming marathon.
No technology or company can survive without keeping up with the pace of innovation in the relevant industry and without improvement. The same is true for the evolution of AoT that will translate into the same for the IoT industry and its capabilities.
The idea is close to one of the topologies distinguished by Davenport – prescriptive analytics. As he states, prescriptive analytics are those that provide specific recommendations based on predictions, experiments, or optimizations. The optimization is a vital part of the efficient functioning of the world of IoT devices. Just like Google applied one machine to manage the energy consumption of its other machines, the same can be made for a range of other connected systems and devices.
An interesting example by Davenport with smart cities explains the benefits for human and urban environment, In some IoT environments, such as smart cities, analytics will need to provide automated prescriptive action. It’s useful to look at a dashboard and know which streets are congested in Singapore, for example, but the real value comes when a system can change traffic light durations and block off freeway entrances based on IoT data.