The Supply Chain Crisis: An Historical Perspective - Edgar Zapata
By Edgar Zapata
Need an excuse? Just say, "it's the supply chain." What first was an understandable shortage of personal protective equipment, because who could have predicted a pandemic, is now a shortage everywhere. The attempted, detailed explanations for the toilet paper scarcity at the start of the pandemic are now just arm-waving followed by muttering "supply-chain." Somehow, those two words are supposed to suffice. We once had Zeus, the fates, fairies, or divine intervention. Now we have something even better to blame when things go amiss, the supply chain.
I can't help but recall the Space Shuttle and its supply chain crises. It had more than a few, but two supply chain crises especially stand out. The first of these crises was underway by the late 1990s, and for me this hit close to home.
You make a one-of-kind unique part for the Space Shuttle, and it's not profitable? Huh?
Our vendor was in town to check out a large valve assembly, part of the rube-goldberg contraptions that connected the bright orange external tank to the Space Shuttles. I looked forward to the visit, having gotten to know the team well. This visit, they gave me the bad news; their company was pulling out of the Shuttle business. After much cajoling, the best response I got was some vague assertion about how the job was no longer profitable. Of course, this boggled my mind. You make a one-of-kind unique part for the Space Shuttle, and it's not profitable? Huh?
After all, this was an era when making a part for the Space Shuttle meant bragging rights. This was so no matter what you made, even if just a bolt. You could do the mandatory advertisement in all the trade magazines, the Space Shuttle lifting off the pad, your company logo, with a very small picture of your product as almost an afterthought. If your product spun at 82,800 rpm, or jammed 75,000 horsepower into a pump that fits in a duffle bag, or some other technology magic (for the times), without coming apart, all the better. You had a right to proudly celebrate being a part of the world's most complex and capable flying machine ever.
Yet somehow, this was no longer profitable?
It's no secret, Space Shuttle contracts covered all costs. They covered even the cost of surprises. Which is to say your company never suffered from having failed to predict all the costs that might arise. Then the government added in a nice profit on top of all that (as a percent). Even better, if you kept your government over-lords (like me) happy there was an extra award on top of that – think of it as the federal government adding a tip. How could this not be profitable?
It turned out this vendor was not alone in fleeing for the hills. Our parts people had been tracking vendors for years and were seeing the writing on the wall. By the late 1990s vendors were leaving the Shuttle program like rats off a sinking ship. The "before" chart showed the Shuttle's various systems, with all the vendors in the branches. The "after" chart showed vendors crossed off – too many, and more leaving soon.
It was the first time I would hear the blame being laid on this mystery called the supply chain.
Not being one to forget a question after a shrug and some simple answer, I investigated what was really happening. How could not being profitable be the end to the question? Remember, a profit was guaranteed, and a nice tip too! A Ferengi could not be happier – the 300th rule of acquisition is "Always take any cost-plus contract, no questions asked."
Along came a couple of better explanations as I looked further into this – "return on capital", and yes, the "supply chain." It was the first time I would hear the blame being laid on this mystery called the supply chain. (Eventually, I would spend some years delving into supply chain management. That's a few years of my life I will never get back. When I left, it was for what I discovered. This tunnel had more dead ends than paths going forward.)
What the figure – "return on capital" meant was just as much of a mystery. Accountants were having a field day here, but for the Shuttle program the explanation could be boiled down to guaranteed profit, yes, but not enough. Companies wanted more because shareholders expected more.
Translated into English, as a practical matter, a Shuttle vendor would tie up a part of their factory making a widget for the Space Shuttle. Floor space was isolated, comings and goings controlled, with everything inside that fenced area tracked obsessively. It could be the spore drive interface for the Starship Discovery or just a sensor. No matter the seeming importance, all the parts could be called anti-social, avoiding mingling with the rest of the people and parts in the factory.
Undeniably, that closed-off, very controlled part of the factory eventually made less profit than the rest of the plant. This was the late 1990s, so why build another plant or expand, with interest rates at 7 percent? You could just kick out the Shuttle business and get the whole plant humming along at the much higher profitability of the rest of the plant. This was "return on capital" at work, if not for us launching Shuttles.
Our NASA Shuttle Logistics Depot would eventually tackle this first supply chain crisis well (through "in-sourcing," they would say; also, sparing no expense.) But it was not the last supply chain crisis in the Shuttle program.
It turns out vendors didn't like just answering the phone on parts made years (OK, decades) earlier.
Because it's always something, soon after, as the space station program was finishing most hardware production, if not yet launched, upper management (some in today's leadership) was made aware that even more vendors were fleeing. This time it was for a new reason. So if it was bad before, it was going to get worse.
It turns out vendors didn't like just answering the phone on parts made years (OK, decades) earlier. The value to a vendor was in the making, not the supporting. The loss of Columbia the next year would morph this problem into yet a worse supply chain crisis – how to maintain a supply chain for whatever came next after the Shuttle when it wasn't clear just what was next. It's hard for a buyer to keep vendors around when you are unsure what you will be buying. Those thousands of vendors – those are links in your supply chain.
To make matters even more complicated, this went beyond the Shuttle. It turned out that small NASA programs and companies with no big shift ahead were fine as long as they could depend on big programs for parts. If a big program went away, well, think of those little fish that follow around a Sharks mouth. No shark? Big problem for little fish.
I eventually left the work on the supply chain management. The field has lots of measures, two of which were efficiency and robustness. (The actual lingo would go into utilization for this, up-side or down-side flexibility for that.) It didn't take long to see the problem.
A very efficient supply chain cannot be very robust, or vice-versa. Yet the illusion modern supply chain management wanted to project was "you can have it all." With the right software, or standards, or model, everything you might need was just a click away.
In retrospect, the Shuttle program's supply chain skirmishes were part of a bigger battle happening across the private sector.
Imagine an engineer spending his days figuring how to get the plant near 100% utilization. Anything less is wasteful. Modern supply chain management software keeps you near this hairy edge. When the surprise comes that no one could have predicted, say you need 10 more valves for the apocalypse, you are not getting those valves. Not any time soon anyway. You are too efficient to be robust.
We were drawing a trend-line nicely connecting the pile of mail-order catalogs to the internet to same-day delivery of groceries to the replicator on Discovery. All of a sudden, where is my instant gratification? While it would be too easy to channel Gordon Gekko on greed and all that, a more complex rhyme and reason may also find supply chain crises are increasingly of our own making.
It's easy to say no one can plan for the next crisis, except they seem to keep coming rather predictably.
Today's start-ups in a world with near-zero interest rates won't be thinking about return on capital for some time. Supply chains are nonetheless being dismantled and reassembled as we speak, courtesy of a pandemic. It's always easier to destroy than to create, to fire than to rehire, and along the way to forget rather than remember. Perhaps your company's rocket engines will come from some other company, or you think this is a good time to just print it all in-house. As has happened before, large NASA programs will complete development and production blocks will follow.
Smaller companies will benefit from these supply chains around them in ways they may not recognize immediately. In all this, rethinking what we value in a supply chain, especially people, and how we balance it all is not a new idea.
While seeming an impossible ask, it is still the ask to have it all. A robust supply chain has margin, valuing people, for a capacity to respond to unexpectedly increased demand. Yet this capacity is too often erroneously seen as waste. This was one takeaway from my foray into this field years ago. Robustness is not a defect, sub-optimization, or over-capacity, the same way reusability is not poor mass management; it's a feature. It's easy to say no one can plan for the next crisis, except they seem to keep coming rather predictably. In all this, perhaps the lesson is "robustness, for lack of a better word, is good."
(Edgar Zapata is a retired NASA engineer and is now a consultant to the space industry. This article originally appeared on his blog zapatatalksnasa.com, and is reproduced with permission.)