This week The Economist put together a special section about how society is overflowing with data. Reading it, I thought it was irony at its best. What better way to cover our data-overload than with a long, drawn-out story bubbling over with … data.
The most obvious example of data-overload is the Internet. When I went to college, during the pre-Internet era, I took an entire class on how to find things in the library. Library computer databases then were in zygote form. They were a beginning, but their breadth was microscopic compared to what the Internet was to become.
In manufacturing, classic examples of data overload come from machine data collection. We now have sensors that show how a vibration may predicate a breakdown of a certain mechanical system. Such data gathering forms the foundation of predictive maintenance (PM), overall equipment effectiveness (OEE), and loads of other acronyms. Do manufacturers use all this data? Sometimes they do, but sometimes the data sit forever in old file directories, just waiting to be deleted.
Enterprise resource planning (ERP) software, manufacturing execution systems (MES), materials resource planning (MRP), customer relationship management (CRM), and other things floating in the alphabet soup of modern manufacturing have changed business. Forecasts still are a murky science—not many predicted the extent of the last recession, for instance—but data collection now allows businesses to predict customer behavior down the road better than ever.
In fact, data collection itself seems to be a good business. SAS, the business software giant, topped Fortune magazine’s list of the best companies to work for in America. The company’s 300-acre campus outside Raleigh, N.C., has on-site health care, massage therapist, flexible work hours, child daycare, and a cafeteria serving free food, among numerous other perks. According to the magazine, SAS CEO Jim Goodnight says he’s not just being a nice guy; it’s good business practice to keep employees happy. That may be true, but the fact that the data business is very profitable may help pay for those perks.
So when does data-gathering become overkill? It probably depends on the type of business. A product-line manufacturer and a contract fab shop have different needs.
One job shop I visited earlier this year measured nearly everything: employee skills, arc-on time in the welding booth, productivity in the laser cutting cells, chemistries used in the powder coat area, and so forth. It’s evidently working for the company, because it doubled its business within the past six months.
A heavy equipment OEM I visited last year had a different philosophy. Workers continually made an effort to develop ever-more-detailed work processes. The better those work processes, the more consistent the company’s manufacturing became. Those procedures included a detailed quality check within each work cell, but the company didn’t have a comprehensive business software package to track every part as it went through the facility. It evidently worked for them. It now takes seven days for raw material to make its way through the plant and into the finished product; and they churn out a machine every three days. That’s light-years ahead of the multi-week (or sometimes multi-month) lead-times the company used to have.
Regardless of the approach, data collection really has become the keystone for modern manufacturing. How a shop collects data depends on the business, and my guess is the method of collection probably doesn’t matter anyway. What really counts is how a company uses the data to shorten the production cycle and quicken response to forever-changing customer demand—no on-site massage therapist required.
I have to admit, though, that a massage therapist on the fab shop floor would be a sight to see.