I read an alarming statistic in a recent Gartner report. It stated that, on average, 70% of factory data goes unused. This was a self-reported metric provided to Gartner from their clients – not an implied or extrapolated number. At first blush, this finding was a bit shocking.
Upon further reflection, my interpretation is that much data is now collected as part of how processes are performed. There may not be much value in this data, so it is simply discarded as part of the process. In this context, the figure of 70% starts to make sense. But, if that much data is being generated and not used, the next question is could it? What better decision support might be possible if just some of that data were better utilized?
The Quest for Better Manufacturing Intelligence
Clearly, the desire to work “smarter” is a popular topic today – and with good reason. New technologies and opportunities now exist to better collect and analyze big data, which can then be presented in an easy-to-understand format that can be acted upon quickly.
This trend helps explain the interest in the Industrial Internet of Things (IIoT). Think of the IIoT as an infrastructure to ease data collection. Investing in this technological transformation is great, but only if you have a complementing strategy to better use that data, once you have it. Here is where the need becomes clear for a data strategy on how to not only increase your ability to store it, but to also access it with greater speed and efficiency.
Those manufacturers that have invested in Enterprise Manufacturing Intelligence (EMI) solutions can recognize a good return on investment. EMI can be used to accomplish this goal provided that the collection mechanisms can work in tandem with the growing needs for intelligence extracted from this source. “Dirty” data or information that must be cleaned, parsed or reconciled before intelligence can be extracted won’t do. Speed of access to support real-time decisions will likely be the differentiating factor in reducing the data that is simply wasted or thrown out vs. that data which becomes useful. Only with this transformation will the 70% figure reported by Gartner decline.
Avoid the Silo Trap
Silos are a trap that has plagued manufacturers for years. Plants operating autonomously without adherence to corporate best practices can hurt overall productivity or continuous improvement programs. Departments that operate in isolation don’t add much value to complementary operations processes. And, in the same way, data silos will also impact your data collection and analysis strategy. Data sitting on “islands” scattered across your organization will not likely be of much help with decision support. It certainly won’t help you to obtain broader visibility into intelligence that enables future predictive analytics.
Given the strategic importance of best capturing and leveraging the right data to improve operational excellence, it would seem that today’s manufacturing organizations should really have a data strategy – one that literally provides the foundation to perform the company’s future growth into new markets or product lines. Yet I don’t hear of too many announcements along these lines … perhaps data strategies are buried in IT initiatives or architecture strategies. Is this the right place? What do you think? I am interested to hear from our readers.
If you liked this article, here are others you might also find interesting:
