The real-time enterprise is increasingly a reality, created as transformative computing trends, including the internet of things (IoT), machine learning, artificial intelligence, and virtualization, redefine what was once the edge of the traditional hub-and-spoke networks. The systems that comprise networks and the devices that tie into them are not only generating, but can also share, a virtually unlimited barrage of streaming data. These systems include everything from hardware, to sensors and mobile phones, as well as software, applications, and databases.
Unlike traditional batch data, streaming data can be continual and immediate—providing enterprises with the ability to act on it as it originates. What’s more, with the right infrastructure in place, this information can be routed to a data lake from which modern business intelligence solutions enable organizations and their disparate business functions to derive actionable insights.
While every organization does not need true real-time data—near real-time will work for most and batch operations are still useful in many settings—all organizations need to consider how they can best manage the increasing flow of data regardless of where or when it occurs. This enterprise information management (EIM) strategy can be developed through close consideration of five key steps:
Define/identify business objectives and ask if real-time data is needed: The drive to embrace the use of data in real-time should be the natural result of business needs that call for immediate access to, or use of, information as it is created. While the use cases are innumerable, real-time applications of data by their nature require a much higher level of network resources than data that is sent every hour or every day, as batch processes often are. Consider this: do you need sensor data from a well head that monitors pressure, status, and flow immediately? Or is once per hour sufficient? Organizations must first consider how frequently information is needed and then set the strategy.
Manage devices at the edge: Advancements in integration, messaging software, and IoT are virtually eliminating the once hard and fast edge of the network. Mobile devices in the modern context can be virtually anywhere, such as a delivery truck that is equipped with a GPS sensor to a mobile phone, tablet, or microchip that monitors the temperature of shipping containers. In cold chain shipping, managing consistent temperatures is just as important as when materials are delivered. Regardless, to be effective organizations need a data and device strategy to ensure that they can “read” the data they need, when they need it. It is also imperative to build asset management strategies for these devices to ensure that the information on them is controlled, secured, and properly maintained. Another increasingly common example of device management at the edge is in healthcare, where tablets and mobile phones are increasingly used at the point of care.
Consolidate your data effectively for use organization wide: As organizations gather and use different kinds of often completely unrelated data forms, this requires the creation of a data lake. Whether this is required goes back to the context of use and the business objective, but in all cases, it is crucial to ensure that a strategy and plan is in place to consolidate the data and store, protect, and back it up.
Consider how the data will be consumed at the point of work: Information for the sake of information can become more of a distraction. Real-time data is no exception. Having data immediately is no guarantee that it will help you achieve your business objectives. Again, it is critical that its use be considered in the development of strategy. Let’s use monitoring again as an example. Do users need to know what is happening all of the time, or just when something is wrong? If they only need to know when something is wrong, what is the best way to relay that information—an alert, a dashboard that’s color coded? The possibilities are limitless, but should reflect a keen understanding of how the information will be used when needed most.
Build in the ability to conduct analytics on-demand: By its very nature, information is not static, it’s dynamic, and so are the use cases that motivate different users to seek and apply it. For many, the information they gain is descriptive, for some it’s diagnostic in nature, and for others it’s predictive. An example can found in the predictive analytics used to proactively identify equipment failure and to guide the resulting maintenance and repairs. For others it is prescriptive and informs what is happening currently to help define what should be happening. Regardless, data’s varied use demands that organizations possess the ability to run analytics in a way that creates actionable, relevant information.
So, why isn’t everyone doing this? The evolution to become a real-time enterprise calls for change in virtually every part of the business. This evolutionary process demands that organizations balance the need to keep the lights on, work on a long-term strategy, and avoid what can rightfully be called the tyranny of the urgent. However, the benefit is worth the investments necessary efforts. A keen focus on the five considerations above will help set organizations on the right path to becoming a real-time enterprise.