"We need more manufacturing jobs in Minnesota" is a common mantra in the media and with politicians.

More jobs are always good and more manufacturing jobs tend be a double win because they are well-paying jobs and they build things — things that we can export from the state to bring back even more money to drive the economy.

But every state wants more manufacturing jobs. We need to think about the problem differently if we are going to get ahead of the pack.

Many of Minnesota's most storied brands are manufacturing companies — think General Mills, 3M, Honeywell, Malt-O-Meal, Rosemount Engineering, Medtronic, St. Jude Medical, Starkey, Arctic Cat, Valspar, Sleep Number, Land O'Lakes, Polaris, Cray, and Control Data to name a few. Many of these brands have been absorbed or left the state, but all of them still support manufacturing jobs here. Understanding why is important to those looking to improve our state's manufacturing fortunes.

Manufacturing has changed. Thirty years ago the companies above were developing new products and new technologies to build those products. They collocated engineering and manufacturing teams because the new products they were creating required new manufacturing processes. Proximity reduced risk and time to market thereby increasing return. Honeywell's ring laser gyroscopes, 3M's traffic safety products, Medtronic's pacemakers, and St. Jude's heart valves were just a few of the new inventions that created new manufacturing processes as well.

But today the face of manufacturing, everywhere not just in Minnesota, is contract manufacturing. Contract manufacturing came about in the early 1980s because high interest rates drove OEMs (original equipment manufacturers) to free their balance sheets of the heavy assets of the factory and focus their capital investments on the product, the intellectual property. The trend accelerated in the 1990s, driven by the internet and its rocket fuel — software.

Contract manufacturers don't own the intellectual property, the secret sauce, of what they build. They replicate manufacturing processes developed by their customers and add value through assembly, test, and supply chain management — all of which are imminently portable. This creates a challenge for economic planners in that the two critical components to their service offering, labor and capital, are both fungible. The availability of labor and portability of capital make it easier to move contract manufacturing and make it less profitable as well. The recent move of 300 manufacturing jobs out of Iowa during a consolidation by Polaris demonstrates the challenge of sustainability in manufacturing today.

But there are exceptions and I have been fortunate to be part of one such example. SkyWater Technology Foundry, formerly Cypress Minnesota Inc., formerly VTC, formerly CDC Microcircuits Division, has been manufacturing semiconductors, perhaps the most sophisticated, competitively manufactured product in history, for over 30 years in Bloomington. Nearly 80 percent of all semiconductors are manufactured in Japan or other parts of Asia today. Moore's Law, the manufacturing axiom that has driven our technology world since 1967, now drives manufacturing growth in Asia. Yet there are still more than 400 operators, maintenance technicians, engineers and administrative professionals manufacturing semiconductor wafers here in Minnesota.

The reason for this, and the reason that Honeywell, 3M, Medtronic, and St. Jude, now part of Abbott Laboratories, all still support manufacturing ecosystems — and that is a key word, ecosystems — here is that these types of manufacturing require both the development and maintenance of technology — not just software, or equipment, or even materials, but full manufacturing process flows. These manufacturing process flows require engineers who continuously develop and maintain the processes used to build new products. They require maintenance technicians and equipment engineers who implement and maintain those processes in a production line that meets global economic pricing competition.

So what should we learn from these exceptions to the trend of migrating manufacturing jobs? Three observations are actionable.

• Manufacturing without engineering, without the creation of intellectual property, is a commodity and hard to defend. The engineering interface to manufacturing empowers the manufacturing leadership to create differentiation and more value for both their customers and themselves.

• Sustainable technology manufacturing requires a wider distribution of technical and vocational education. One of the most critical roles in the SkyWater manufacturing facility is the maintenance technicians who perform deep critical thinking and problem-solving exercises, under pressure, every time a piece of equipment doesn't perform to the necessary specifications.

• Manufacturing technology needs a financial environment, both public and private, that understands global competition at an everyday pricing level as well as strategic differentiation over the long haul. A pure software product development can be complete in 180 days. The rule of thumb for full product developments, hardware and software, is 18 months. A new manufacturing process, e.g. the next DNA-enabled therapy, can take five to 10 years.

Minnesota is fortunate to have a long agricultural and manufacturing tradition. Many in today's business press lament our inability to replicate the speed and financial agility of Silicon Valley. They obsess over the number of start-ups and amount of venture capital. But our technology history is not populated with unicorns.

Our legacy is borne by the long and steady pull of a team of draft breeds — technical, vocational, financial, and operational. Sustainable manufacturing, technology-based manufacturing, requires patience and perseverance — fortunately, traditional Midwestern values.

Scott Nelson is chief technology officer at SkyWater Technology Foundry, Bloomington.