Is Your Data Management Infrastructure Modern Enough for IoT?

July 10, 2018

Is Your Data Management Infrastructure Modern Enough for IoT?

Internet-of-Things (IoT) has entered the lexicon of IT-related buzz terms in a big way over the past few years, and there is good reason for this.  IoT at its foundation refers to what can literally be billions of devices spanning the globe (and beyond) that can be connected to the internet to serve a variety of purposes. Both businesses and consumers can and will reap significant benefits from what IoT has to offer.

IoT is supported by a variety of technologies – computer systems, networks, end user devices, software – but at the heart of IoT is the collection, storage, processing, and analysis of data. Data growth is nothing new, of course. The volume and even types of data relevant to doing business have been growing at an accelerated pace for some time, in great part as an outgrowth of web-based applications designed to reach consumers and support end user activities such as social media.

Most organizations today have built and are using data management infrastructures which were often built on data warehouses used to collect and analyze data. Many have evolved to be more compatible with the influx of unstructured data fed by the web and other sources.

However, because silicon technology has progressed quickly, an assumption was made in many cases that storage is, and would, remain almost infinitely plentiful – and cheap. IoT may force a reconsideration of that assumption, with data from the aforementioned billions of devices taxing storage resources.

What this example leads to is that while organizations will want to leverage their existing data management infrastructures, they will need to modernize them in some significant ways. Some of the factors driving an evaluation of existing data management capabilities and their suitability for IoT include:

  • The ability to scale to the needs of an expansive IoT environment. This problem is not just about “size” (as in, “I need more storage”, or “my servers don’t keep up with the analytics I need”), but also about variety. If, as an example, a manufacturer or other type of business has operations in multiple locations and needs to consolidate them into one IoT environment, there are likely incompatibilities among systems, software, and other technologies that affect the ability to scale.
  • Storage is of course not infinite, so new ways need to be devised to process and analyze data to mitigate the storage issue. While the cloud can provide a less expensive and more flexible storage option in some cases, this is only part of the story. One use of analytics involves the extraction of relevant information required in real time prior to storage. This can not only help reduce long term storage requirements but can provide insight and actionable information faster.  Analytics “at the edge” can also help here.
  • Data is of high value, but only if it is complete and correct. In a business environment, incorrect data can result in lost business and higher operational costs. The amounts of data a highly scaled IoT environment represents raise the stakes here. There is a need therefore for data quality tools that can identify and act on data deficiencies before they affect the business.

IoT leverages technologies and tools, but clearly there is more to it than that. Billions of sensors, the data that flows as a result, and the actionable intelligence that is generated and acted are part of a process that goes way beyond the underlying equipment on which it is built.  In future blogs, I will discuss how organizations can effectively deal with their data modernization issues in more detail.

Steve Garone, Uncategorized , , , ,
About Steve Garone