As you read this, billions of devices, sensors, machines and accessories are being networked to prepare for the Internet of Things (IoT). Gartner predicts that the IoT will include 26 billion installed units by 2020. That will generate a staggering volume of data to store and present huge challenges for information technology to analyze. Many people are now asking how we can glean fast insight from the data. How can we turn the data generated by the IoT into tangible benefits?
'Real-time analytics' is a term that gets used a lot these days. On the surface it seems like the perfect way to quickly derive meaningful insights from the IoT. However, 'real time' usually refers to fast, interactive analysis of static datasets, not live data.
It might be last week’s smart grid telemetry or last month’s manufacturing floor machine logs. While the fast analysis of static data helps identify important data patterns and long term trends, it leaves a critical gap between the identification of a pattern and the use of that intelligence to capture business opportunities in the moment.
The real challenge created by the IoT will be to apply real-time analytics to live, continuously-updated data with extremely low latency. Known as 'operational intelligence,' this concept has been around for years, but until now the technology to realise it has lagged.
Operational intelligence continuously analyses live data and maintains a dynamic model of a real-world system. Unlike conventional approaches which analyse streaming data, this technique integrates live data and historical information to provide a more complete picture of IoT activity. This enables a deeper understanding of customer needs and business opportunities and provides timely, effective feedback.
Although implementing operational intelligence introduces significant challenges in handling large volumes of data with fast response time and highly availability, it creates amazing new possibilities for the IoT – the kind of possibilities that transform industries. Here are some examples followed by a peek at the technology that makes operational intelligence possible.
Predicting the unpredictable
Imagine a commercial food processor that uses machines with built-in sensors capable of streaming continuous updates for vital operating data to an analytics system. Simply analysing historical data greatly reduces the chances of catching a possible failure before it occurs. However, by using operational intelligence, the company can compare streams of live data with historical models to identify early indicators of problems and thereby prevent costly failure scenarios.
This is just one example of how manufacturing can be advanced through real-time analysis of live data. Data from machinery and manufacturing processes will make up a significant percentage of the overall data created by the IoT, and this will create many opportunities for businesses to take full advantage of this intelligence to improve both their processes and their bottom line.
Giving brick-and-mortar retailers an advantage
The growth and maturity of e-commerce poses an increasing threat to traditional brick-and-mortar retailers. While many consumers prefer the experience of picking out items in a store, the convenience of online shopping is undeniable. According to ShopperTrak, during the 2013 holiday shopping season U.S. retailers received approximately half the foot traffic they had experienced just three years prior. With operational intelligence, brick and mortar stores have the opportunity to level the playing field by tracking customers and inventory in real time; inventory has become an IoT for these retailers.
For example, when a consumer walks into a store, she can opt-in to an enhanced shopping experience that integrates her location and preferences with her shopping history. Operational intelligence empowers sales associates to make personalized recommendations for the consumer to better match her needs, thereby combining the personalisation of online shopping with the tactile experience of in-store shopping.
Moreover, this technology enables the store to track inventory with high efficiency using RFID tags, reducing the stock that needs to be kept on hand and ensuring immediate access to requested items. By offering the shopper highly personalized and relevant offers and managing inventory with new levels of efficiency, operational intelligence gives the brick and mortar retailer a distinct competitive edge.
Making cable TV smarter
The spread of cable TV has changed lifestyles by providing an explosive number of entertainment options for the viewer. In fact, there are so many choices that customers sometimes have difficulty in making a selection but at the same time have a healthy appetite for even more options.
Telecommunications providers understand these challenges and are beginning to use operational intelligence to help viewers identify new entertainment options personalised to their tastes. For example, by applying operational intelligence to data streamed from set-top boxes combined with a viewer’s historical viewing patterns, the customer can be alerted to an upcoming program featuring a favorite sports team or to product offers that match both the current show and the viewer’s profile.
The cable provider also can monitor and analyse set-top box data to quickly identify and address issues with viewing quality or network speed while displaying messages to the viewer if needed and quickly dispatching a service representative. These new capabilities both enhance the customer’s experience and offer upsell opportunities for the provider, while lowering operating costs.
The technology behind operational intelligence
Companies have been analysing their historical data for years to create business intelligence, that is, to identify patterns and trends hidden in the data. They have learned about their customers’ preferences and behavior and then applied these learnings to improve their product and service offerings. Operational intelligence enables companies to take the next step so that data analytics can be carried out and acted on in the moment. The value of operational intelligence is clear, but what does it take to make it a reality?
Operational intelligence for the IoT requires a computing environment that can store, rapidly update, and continuously analyse data representing a large population of real-world entities, such as devices, inventory items, and customers. The secret sauce that makes this possible is in-memory computing. This technology has been used for more than a decade to store and update large, fast-changing data sets.
More recently, scalable, in-memory computing capabilities have been introduced to give this computing platform the horsepower needed to rapidly analyse live data and provide immediate feedback. For example this scalable computing technology can analyse a terabyte of continuously changing data in a few seconds or less.
Unlike analytics platforms designed to analyse static data in a back office, in-memory computing platforms can be integrated directly into live systems, reliably tracking and analysing large populations of fast-changing data. When designed to meet the stringent needs for high availability in commercial environments, these platforms enable companies to obtain the operational intelligence they need to manage the IoT. In-memory computing products with integrated, continuous high availability, such as ScaleOut Software’s ScaleOut Analytics Server, are specifically targeted for use in live, mission-critical systems.
Operational intelligence gives companies the means to leverage the fast-changing data created by the IoT, and it empowers them to enhance the customer experience, optimise business processes, and grow profits. With their scalable performance and integrated high availability, in-memory computing platforms now open the door to operational intelligence for systems which manage the IoT. Welcome to the next age of transformation.
Sourced from David Brinker, COO, ScaleOut Software