PR Challenges in Leveraging Big Data for Process Industries

Our Mission at Seeq Is to Accelerate Analytics in Process Industries

December 20, 2015

Our mission at Seeq is to accelerate analytics in process industries. To achieve this, our industrial process experts collaborate with our software team and analytics engineers to identify and leverage the best technologies available. Which led us to build Seeq on technologies and platforms generally known as “big data,” for which we were recognized by the Puget Sound Business Journal (PSBJ). Seeq Corporation was included in PSBJ’s October, 2015 list of top big data software companies in the Puget Sound area.

This is a bit ironic. Of course we appreciate the recognition, and any time we can be on a list in the proximity of innovative firms like Tableau Software (#1), Socrata (#5) and Bizable (#13), we will welcome the comparison.

At the same time, we’ve learned a thing or two about big data’s perception issues in talking with customers, and much of it has not been good. As a result, we now conspicuously avoid the term (try finding it on our web site) and generally shy away from bringing up “how we built Seeq” until we’re asked.

What were the issues and misunderstandings we heard when we used to talk about big data?

First, big data frequently implies a platform approach requiring developer expertise in order to build a solution against the API. This isn’t true of course; using Google (built originally on what became Hadoop) doesn’t require MapReduce skills, voice recognition in Cortana or Siri (built using deep learning) doesn’t require a data scientist, and LinkedIn and Facebook (built on property graph databases) don’t require expertise in NoSQL.

So leveraging or building on big data can result in end-use applications with no programming skills required, which is probably why none of these products ever mention big data either. This was the first mismatch: Seeq is an application. And while it has an API, developer skills are absolutely not required for successful use.

The second issue is that big data isn’t considered by engineers and managers in the process industries to be part of an OT or industrial offering. Instead, many professionals involved in process manufacturing believe that big data is associated with IT organizations and is in the domain of the experts there with the deployment, architecture, and programming skills to successfully implement large, complex projects. Again, this expectation for IT and big data is a mismatch with Seeq, which has employees with 100+ years of expertise in manufacturing data and is squarely focused on bringing new tools to engineers and other operations personnel in process industries.

The third issue is that many big data companies, in addressing the needs of process manufacturing,propose their solution as yet another centralized database – sometimes referred to as a data lake – to which data must be exported. In this environment, data scientists can then apply scripts and algorithms to the data. Actually, applications built on big data should make no such requirement. Google doesn’t import all of the data in the whole of the web, rather it indexes the content for easy reference. Seeq follows this paradigm as well: leave your data where it is and Seeq will make it accessible for analysis without the requirement to store it yet again.

Finally, the last misunderstanding regarding big data is the expectation of a long, long, time to benefit. This is in part due to the assumption of big data as a developer platform, plus the assumption of IT overhead, but more than that it’s simply the overwhelming complexity of the big data ecosystem and its offerings. Any two vendors are likely to disagree on the definition of big data, much less how or why any single offering is better than any other and… it’s just so complicated. For customers who have issues right now, the prospect of a distant and convoluted path to insight, looping in other groups and requiring new skill sets, comes across as a mismatch to their needs. And again, Seeq’s ability to get installed, connected to data, and ready to use in minutes is not in line with expectations of a long time to big data benefits.

So as we met with customers earlier this year and last, the feedback was clear to us: while big data technologies were driving Seeq innovation and features for customers, that fact was best left off our discussion list. Big data may be the “how” of Seeq capabilities, but given its overhead in misunderstanding and negative expectations, it’s not something to focus on.

Thanks again to the Puget Sound Business Journal for the recognition, but going forward we’ll be keeping a low profile in the big data category.