Shipulski On Design

Archive for December, 2011

There’s an unnatural attraction to lean – a methodology to change the value stream to reduce waste.  And it’s the same with Design for Manufacturing (DFM) – a methodology to design out cost of your piece-parts. The real rain maker is Design for Assembly (DFA) which eliminates parts altogether (50% reductions are commonplace.) DFA is far more powerful.

The cost for a designed out part is zero.  Floor space for a designed out part is zero. Transportation cost for a designed out part is zero. (Can you say Green?) From a lean perspective, for a designed out part there is zero waste.  For a designed out part the seven wastes do not apply.

Here’s a recipe for unreasonable profits:

Design out half the parts with DFA.  For the ones that remain, choose the three highest cost parts and design out the cost.  Then, and only then, do lean on the manufacturing processes.

For a video version of the post, see this link: (Video embedded below.)

A Recipe for Unreasonable Profits.

It’s been a long slide from greatness for US manufacturing.  It’s been downhill since the 70s – a multi-decade slide.  Lately there’s a lot of hype about a manufacturing renaissance in the US – re-shoring, on-shoring, right-shoring.  But the celebration misguided.  A real, sustainable return to greatness will take decades, decades of single-minded focus, coordination, alignment and hard work – industry, government, and academia in it together for the long haul.

To return to greatness, the number of new manufacturing jobs to be created is distressing. 100,000 new manufacturing jobs is paltry. And today there is a severe skills gap.  Today there are unfilled manufacturing jobs because there’s no one to do the work. No one has the skills. With so many without jobs it sad.  No, it’s a shame.  And the manufacturing talent pipeline is dry – priming before filling.  Creating a sea of new manufacturing jobs will be hard, but filling them will be harder.  What can we do?

The first thing to do is make list of all the open manufacturing jobs and categorize them. Sort them by themes: by discipline, skills, experience, tools.  Use the themes to create training programs, train people, and fill the open jobs. (Demonstrate coordinated work of government, industry, and academia.)  Then, using the learning, repeat.  Define themes of open manufacturing jobs, create training programs, train, and fill the jobs.  After doing this several times there will be sufficient knowledge to predict needed skills and proactive training can begin.  This cycle should continue for decades.

Now the tough parts – transcending our short time horizon and finding the money.  Our time horizon is limited to the presidential election cycle – four years, but the manufacturing rebirth will take decades. Our four year time horizon prevents success. There needs to be a guiding force that maintains consistency of purpose – manufacturing resurgence – a consistency of purpose for decades.  And the resurgence cannot require additional money. (There isn’t any.)  So who has a long time horizon and money?

The DoD has both – the long term view (the military is not elected or appointed) and the money.  (They buy a lot of stuff.) Before you call me a war hawk, this is simply a marriage of convenience.  I wish there was, but there is no better option.

The DoD should pull together their biggest contractors (industry) and decree that the stuff they buy will have radically reduced cost signatures and teach them and their sub-tier folks how to get it done.  No cost reduction, no contract.  (There’s no reason military stuff should cost what it does, other than the DoD contractors don’t know how design things cost effectively.) The DoD should educate their contractors how to design products to reduce material cost, assembly time, supply chain complexity, and time to market and demand the suppliers.  Then, demand they demonstrate the learning by designing the next generation stuff.  (We mistakenly limit manufacturing to making, when, in fact, radical improvement is realized when we see manufacturing as designing and making.)

The DoD should increase its applied research at the expense of its basic research.  They should fund applied research that solves real problems that result in reduced cost signatures, reduce total cost of ownership, and improved performance.  Likely, they should fund technologies to improve engineering tools, technologies that make themselves energy independent and new materials.  Once used in production-grade systems, the new technologies will spill into non-DoD world (broad industry application) and create new generation products and a sea of manufacturing jobs.

I think this is approach has a balanced time horizon – fill manufacturing jobs now and do the long term work to create millions of manufacturing jobs in the future.

Yes, the DoD is at the center of the approach. Yes, some have a problem with that.  Yes, it’s a marriage of convenience. Yes, it requires coordination among DoD, industry, and academia.  Yes, that’s almost impossible to imagine. Yes, it requires consistency of purpose over decades. And, yes, it’s the best way I know.

Design for Manufacturing (DFM) is all about reducing the cost of piece-parts. Design for Assembly is all about reducing the cost of putting things together (assembly).  What’s often forgotten is that function comes first.  Change the design to reduce part cost, but make sure the product functions well.  Change the parts (eliminate them) to reduce assembly cost, but make sure the product functions well.

Paradoxically, DFM and DFA are all about function.

Here’s a link to a short video that explains DFM and DFA: link to video. (and embedded below)

Although it’s more important than ever, trust is in short supply. With everyone doing three jobs, there’s really no time for anything but a trust-based approach. Yet we’re blocked by the fear that trust means loss of control.  But that’s backward.

Trust is a funny thing.  If you have it, you don’t need it.  If you don’t have it, you need it. If you have it, it’s clear what to do – just behave like you should be trusted. If you don’t have it, it’s less clear what to do. But you should do the same thing – behave like you should be trusted.  Either way, whether you have it or not, behave like you should be trusted.

Trust is only given after you’ve behaved like you should be trusted. It’s paid in arrears. And people that should be trusted make choices.  Whether it’s an approach, a methodology, a technology, or a design, they choose.  People that should be trusted make decisions with incomplete data and have a bias for action.  They figure out the right thing to do, then do it.  Then they present results – in arrears.

I can’t choose – I don’t have permission. To that I say you’ve chosen not to choose. Of course you don’t have permission.  Like trust, it’s paid in arrears.  You don’t get permission until you demonstrate you don’t need it.  If you had permission, the work would not be worth your time. You should do the work you should have permission to do.  No permission is the same as no trust.  Restating, I can’t choose – I don’t have trust. To that I say you’ve chosen not to choose.

There’s a misperception that minimizing trust minimizes risk. With our control mechanisms we try to design out reliance on trust – standardized templates, standardized process, consensus-based decision making. But it always comes down to trust.  In the end, the subject matter experts decide. They decide how to fill out the templates, decided how to follow the process, and decide how consensus decisions are made. The subject matter experts choose the technical approach, the topology, the materials and geometries, and the design details. Maybe not the what, but they certainly choose the how.

Instead of trying to control, it’s more effective to trust up front – to acknowledge and behave like trust is always part of the equation.  With trust there is less bureaucracy, less overhead, more productivity, better work, and even magic.  With trust there is a personal connection to the work.  With trust there is engagement.  And with trust there is more control.

But it’s not really control.  When subject matter experts are trusted, they seek input from project leaders.  They know their input has value so they ask for context and make decisions that fit.  Instead of a herd of cats, they’re a swarm of bees. Paradoxically, with a trust-based approach you amplify the good parts of control without the control parts.  It’s better than control. It’s where ideas, thoughts and feelings are shared openly and respectfully; it’s where there’s learning through disagreement; it’s where the best business decisions are made; it’s where trust is the foundation.  It’s a trust-based approach.

The No. 1 reason initiatives are successful is support from top management. Whether it’s lean, Six Sigma, Design for Six Sigma or any program, top management support is vital. No argument. It’s easy with full support, but there’s never enough to go around.

But that’s the way it should be. Top management has a lot going on, much of it we don’t see: legal matters, business relationships, press conferences, the company’s future. If all programs had top management support, they would fail due to resource cannibalization. And we’d have real fodder for our favorite complaint—too many managers.

When there’s insufficient top management support, we have a choice. We can look outside and play the blame game. “This company doesn’t do anything right.” Or we can look inside and choose how we’ll do our work. It’s easy to look outside, then fabricate excuses to do nothing. It’s difficult to look inside, then create the future, especially when we’re drowning in the now. Layer on a new initiative, and frustration is natural. But it’s a choice.

We will always have more work than time….

Link to complete article