When I graduated with an engineering degree some decades ago, I learned that the organizations I was going to work for had internal communication problems. This was especially true for those that designed and manufactured complex machinery such as engines, aircraft, or automobiles. Design departments designed things, manufacturing departments manufactured them, quality departments measured them, and service departments serviced them—but rarely did they talk to each other. As a result, it was hard to manufacture products, harder to service them, and recalls were rampant. Quality was an issue.
We had names for these problems. Designers “threw designs over the wall,” or organizations created “stovepipes.” But that was the 1980s. I had thought that new technology and simple human awareness had solved many of these issues. Apparently not.
This was pointed out recently by Norbert Hanke, president of Hexagon Manufacturing Intelligence, speaking in June 2018. He noted that even today, stovepipes and throwing designs over the wall are a problem. He was speaking in the context of metrology equipment and GD&T as it is embedded in today’s smart factories. He is not alone, and these views require some explanation.
First, progress did not stall. Even without employing technology, designers and shop-floor engineers talk more than in the 1980s dark ages.
However, the 1980s problems have grown. Products are infinitely more complex. Customers demand exceptional quality. Mechanical tolerances are tighter, and shapes are more complex. As factories are getting smarter, integrating a complex product design into them smartly is equally complex. Stovepipes, more related to data than human communication, have crept back.
Solving these more complex problems is the heart of some concepts around Industry 4.0 and an even newer term, Quality 4.0. As Nikon’s Corporate Vice President Tadashi Nakayama explained in a recent press release, “the core elements of a Quality 4.0-based process require automated measurement of key features on components as close to real-time as possible, plus the need to acquire digital results and feed them back directly to machinery to control production automatically.”
This is one area where the problem has grown. The level of automation and integration between machinery and metrology Nakayama described will require a higher level of communication in terms of data—especially between those who design and those who measure products. Achieving “automated measurement of key features” will require CAD models with attached GD&T, sometimes known as Model Based Design (MBD). A key trend to watch here is how valid is that attached GD&T. If the designer does not do a good job, forget automated measurement.
Another dimension is that Design for Manufacturing and Assembly is still important. There must be better two-way communication. Fortunately, product designers have access to generative design techniques and multi-domain optimization tools.
Using optimization techniques will require simulation tools as well. Clearly, this is what Hexagon Manufacturing Intelligence seems to think—it has been expanding into this area by acquiring CAE software and other technologies. It has a comprehensive lineup of software technologies, from CMMs to finite element software.
In my view, all roads to eliminating 21st-century stovepipes lead to the product designers. It will be especially important to educate people who create CAD models and derive and attach GD&T to them, while accepting and using manufacturing constraints in creating them.