When I talk with my customers about their manufacturing data, they don’t often call it “Big Data.” From what I can tell, they could care less about labels and trends in the IT world. They’re far more interested in how to turn their silos of discrete data into actionable information, and how they can do it NOW.
Those are the three big issues:
- Having silos of data
- Lacking a method for turning data into action, and
- The need for data NOW.
There is no shortage of data in modern manufacturing companies. The problem is that much of this data is stuck in special purpose silos. It isn’t uncommon to see companies with dozens if not hundreds of these silos.
In our urban world, the term “silo” seems archaic. Where I live, a small city on the edge of one of the largest Amish communities in North America, silos are a fact of life. Drive a few minutes from my office and you’ll find the country dotted with small family farms, most with 50′ or 100′ tall silos storing what I presume is a year’s supply of animal feed. In this context, silos serve a singular and useful purpose: to store lots of feed.
Surely, the silos of data in the manufacturing world also serve a singular and useful purpose – otherwise why were they built in the first place? But their “silo-ness” is their great limitation. Storing piles of data in a silo sub-optimizes an organization. Silos block our view and appreciation of the entire system. We have to create links between these silos so people can see the vital connections between the parts of the system.
Breaking down the silos begins to address the second big issue: by what method do we turn data into action?
The appreciation of a system was W. Edwards Deming’s second of four components of his system of profound knowledge. If we can break down silos of data, we’re better able to appreciate the system.
But appreciating the system isn’t enough. Deming understood this, and a second key component of his system of profound knowledge was the knowledge of variation.
We use knowledge of variation to understand whether the variation in the system is significant. Control charts and other statistical tools are based on a reliable theory of variation, and they lay the foundation for turning data into knowledge.
The value of NOW
If we’ve broken down the silos so we can see the interrelated components of our system, and we apply a dependable theory of variation to our data, the final issue of dealing with Big Data in manufacturing is the time it takes to get from question to answer. Put another way, if we have a problem, how soon do we know about it so we can take corrective action?
For much of the history of manufacturing data systems, control of the data was in the hands of a few. The Big Data revolution has helped democratize access to data because systems can cut out the middleman and give direct access to all key stakeholders.
Cutting out the middleman collapses the cycle time for using the data. In the process of cutting out the middleman, we cut the cycle time from curiosity to answer. We cut the time it takes to react to and fix problems.
Now users can readily put their hands on the data and take immediate corrective action.
In the manufacturing world, access to real-time data is critical because so much is at stake over such short periods of time. When the data tells you —in real time— when and what you need to pay attention to, the value of the data goes through the roof.
We see this occurring all the time. You can read several examples of this here.
What about you? How does real-time data drive value for your business? What does not having it cost you? Email me at ejmiller [at] hertzler [dot] com, or post a comment below. I’d love to hear your perspective.
