Without Things, There Is No Analytics Of Things (AoT)

I was recently having a discussion with Richard Hackathorn, an industry strategist & analyst and Dan Graham, a colleague who is deep into my company’s strategy for the Internet of Things (IoT). We were specifically talking about how to enable the Analytics of Things (AoT) and what barriers and opportunities exist today.

During the discussion, it hit me that one of the biggest hurdles faced in trying to purse the AoT is actually just a new iteration of a common, recurring problem that has vexed analytics professionals for years. Namely, we can’t analyze anything related to IoT until the infrastructure investment is made to create, acquire, and make available the data necessary for analysis. I recently wrote about one key aspect of the infrastructure related to automating the tracking and management of all of our things. Let’s address the need for it here.

… impossible! This is obvious, of course. But, it is a very critical point both today and in the past. Early in my career, we had all sorts of wonderful ideas about how to analyze data. However, the data itself was incredibly difficult, if not impossible, to access for analytical purposes. The move from mainframes to enterprise analytic platforms certainly helped. But, to use those platforms to advantage required that the platforms made available the data we actually needed for our analysis.

Unfortunately, it often took years and a lot of pleading and business case building to convince an organization to invest the resources required to enable our analytic vision. Just when we thought we had all we needed, another new set of data would enter the scene and start the entire process all over again. It was worth it, however, because time has proven that new data improves the power of analytics immensely. The world of big data has recently reaffirmed this, and we can expect Internet of Things data to also reaffirm it.

That brings us to the current need to begin exploring the AoT. There are plenty of wonderful ideas and plans out there for the AoT. However, the biggest barrier for most organizations is actually getting the data required in the hands of those who can analyze it. There are several reasons that this is difficult today:

Clearly, the AoT can be nothing but a dream until an organization takes care of the first two steps. Those first steps also require not only a large investment, but a timeline based on quarters, if not years, instead of weeks and months. In the short term, analysts can acquire inexpensive sensors and begin working with the data to gain an understanding of how IoT data streams work. For example, I have a colleague who outfitted his boat with dozens of sensors and now explores that data for fun and practice. When robust corporate IoT data is available, he’ll be ready to hit the ground running.