On a quest to reach the pinnacle of quality control, manufacturers in the automotive, industrial equipment and other industries are collecting and crunching massive amounts of information to best understand and “see” exactly what is happening on the plant floor. Serving as their eyes in this scenario is manufacturing intelligence software coupled with big data. Together, these technologies can provide a detailed view of everything in the work cell—right down to the nuts and bolts, wheels and drivetrains.
More and more, manufacturers are seeking granular visibility, as outlined in this Wall Street Journal article, which highlights how Raytheon and Harley-Davidson are using MES software and automated data collection techniques to track everything from the humidity in the room to the turn of a screw. Randy Stevenson, a missile-systems executive at Raytheon, is quoted in the WSJ article as saying, “If a screw is supposed to be turned 13 times after it is inserted but is instead turned only 12 times, an error message flashes and production of the missile or component halts.” Improvising with a defective screw or the wrong size screw isn’t an option, he says. “It’s either right or it’s not.”
Whether it’s a missile, a motorcycle, or a milk carton, quality is of the utmost importance. And, there are a “factory of things” out there, that, no matter how minor they seem in the making of a product, could potentially cause major problems somewhere in the product lifecycle if not proactively and accurately monitored and controlled. That’s why Harley-Davidson keeps a continuous record of the tiniest production details right down to the speed of the fans in the painting booth. Now everything – including the paint job – is considered an exact science. There is no room for artistic leeway.
Seeing Through a Sea of Data
Where to find the data and how to measure it, however, is a bit like chasing a wave when it comes to global manufacturing operations. Data is seldom static and seems to always be on the move – from the instrument to the controller to the switch to the server – there is always data that is changing and flowing through the network.
In a recent AutomationWorld column, Gary Mintchell explores the impact of data within the test and measurement world, noting that there are traditionally four variables of data:
Recently, the term “Visibility” has been added to the mix, as a 5th “V.” That’s because manufacturers need to move beyond simply tracking how much and what kind of data they have, but need to start focusing more on how they are actually seeing the data – and how current is it – in order to start analyzing the results.
New big data solutions, born from partnerships between Manufacturing Execution Systems (MES), Manufacturing Intelligence (MI), and Big Data vendors are capturing information from the many “things” scattered inside and outside of the plant. Mintchell calls this emerging trend “big analog data,” with the focus being on the ability to process terabytes of streaming data originating from thousands of real-time sources.
When it comes to the future of predictive analytics and quality control, it is no longer about measuring specific key performance indicators (KPIs). Rather, it is about measuring everything in order to create complete transparency across global production facilities as well as throughout the supply chain and distribution pipelines.
Ultimately, plant manager and CEOs alike should always have an answer to the question, “What do you see?”