Rosie the Riveter vs the US NAVY

Or 

Better vs Good Enough

Want a new idea? Read an old book… I was reading about Liberty ships…

During World War II US shipyards built 2,710 Liberty ships. The Liberty ship started out as the “Ocean” class cargo ship, a British design (some 60 were built in various British yards). But US shipbuilders, led by Henry Kaiser, studied then changed the design, simplifying and standardizing it. One major change was that the US ships would be welded (the British ships were riveted); welds were not only stronger, but a faster production process. More welders would be needed but they would be trained. 

As an aside, within a year many welders were drafted, and the welders were replaced by woman, who were trained how to weld, and construction continued. This is worthy of note as we continue to hear that there aren't enough skilled workers to man our current shipyards - which are turning out just a handful of ships each year, yet more welders and other skilled workers can't be trained fast enough? Hmmmm…

A number of years ago I had the odd opportunity to hear a 4-star admiral intellectually rough up another 4-star and 3 x 3-star admirals. The subject was aircraft carriers, specifically, why they were so hard to build and even more, why they were so hard to refuel. The admiral pointed out that the Navy keeps changing the ships - each one is a bit different. And those changes add up to, as he said, each ship being, in effect, a separate class, a unique hull. So, when it comes time to do something major, we find ourselves “re-inventing” the process.

This is true in much the Navy does, and the other services as well. Every weapon system has any number of versions. Some of the versions are minor, some are major. But they are different. Software loads in particular these days make for, in some cases significant changes between one weapon system and another. This is justified because: “Were incorporating lessons learned.” Consider, for example, the F-18. First there was the F-18A. Then the F-18B, which was a two seat version of the A, Then the F-18C/D which were significant upgrades of the A and B.

F-18Es (single seat) and F-18Fs (two seat) look like earlier F-18s but are different airframes, different engines, and different avionics, and are significantly more capable. And there are variants of the F-18E/F.

But the differences don’t stop there. Software and hardware changes result in designations that tell of different capabilities: F-18C block 10, block 20… block 50… some of which involve substantial changes (new radars). Not that the US Navy is alone in this; consider the USAF F-16, of which I believe there are 12 or 13 variants.

Each variant is justified as a necessary improvement in one or more capabilities. But with change often come production and maintenance complexities and additional training requirements. These can be justified up to a point. But it’s incumbent upon leadership to judge when enough is enough, and accept risk in favor of more rapid construction, lower costs, etc. In short, leadership needs to know when “better” is the enemy of “good enough.”

In the recent past, shipbuilding turned into a bone in the throat to the US Navy. It was not always so. A review of Navy procurement during WWII reveals more than simply “we built a lot of ships.” We did. But we also became very efficient at doing so.

Consider the Fletcher class destroyer, one of the most successful combatants ever made. 175 Fletcher class destroyers were made between 1941 and 1945, in 11 different shipyards. Design began in the autumn of 1939 - after the war started in Europe, with the lead unit laid down in October 1941, commissioned in June 1942 - 9 months. She was in the Solomon’s - at war - by October of 42. The last ship was completed in 1945. The Navy kept many on active duty more 25 years, many were transferred to allied navies, the last one was retired (from the Mexican Navy) in 2001.

It’s worth noting that during the 1930s Navy planners had worked out, in detail, the type and number of ships that would be needed to execute US war plans, plans for a war that would include Germany and Japan. In the case of the Navy, the design criteria for fighting a war in the Pacific - tremendous range requirements - was clear. Other requirements included huge amounts of lift (cargo ships) and amphibious capabilities, and the huge organization that would be required to establish sea control (sink submarines, defeat other naval forces, seize ports, etc.) and the scores of different capabilities such a navy would need. 

So, lots of ships needed to be built. Again, the Liberty ship is instructive: 18 US shipyards made 2,710 Liberty ships. As with the Fletcher DD, multiple yards turned out the same product. Engineers studied the Ocean class design and spent time simplifying and standardizing the design, so that each ship was a copy of the last one (or nearly so). The ships were built in pieces and then the pieces assembled; by 1943 the average assumably time had been reduced to 24 days - from keel laying to launching. It still took almost 600,000 man-hours to build the ship (the overall fleet average was a 592,000 man-hours per ship), but it was possible to assemble one in less than a month.

These 2700 ships were followed by 531 Victory ships. Meanwhile, we built more than 30 large attack carriers. More than 100 light and escort carriers. 10 Battleships. Scores of heavy cruisers. Perhaps 100 light cruisers. Hundreds and hundreds of destroyers and destroyer escorts. Hundreds of submarines. Thousands of amphibious warfare ships. And virtually all of them welded together. Yet the yards were able to train up sufficient welders to sustain these phenomenal production rates. One might argue that while “Rosie the Riveter” is mnemonically catchy, Rosie was probably welding, not riveting, but “Wendy the Welder” doesn’t resonate.

To put this in a modern perspective, per on-line point papers to Congress, a Virginia class submarine requires some 5 million man-hours to build. Much of that is not at the shipyard, there are literally thousands and thousands of sub contractors - spread across all 50 states - who participate in the construction of these subs. 

But 5 million man-hours is the output of 2,500 workers each year. If Congress could add 2,500 workers to the submarine production process, that would equate to the equivalent of 1 more hull - equivalent - produced per year. These are not huge numbers of workers.

What has all this got to do with today’s Navy?

First, have the shipyards sat down with the Navy and worked out ways to simplify designs? Maybe that needs to be done again. Take a hard turn on the designs.

Second, can we take a more risk and freeze designs and modifications until we get basic processes to work and get production numbers increased? What is “good enough?”

Third, fix the skilled worker problem - it’s been done before.

Fourth, put someone in charge. Reports directly to SecNav. Or SecWar. No other portfolio, just ship construction. No office in the Pentagon. Maybe an office in Newport News, another in Bath, Maine, another in Pascagoula, another…you get the point. Cancel leaves, daily meetings. Have the offices uncomfortably close to the managers of the shipyards… they should be tripping over each other… give him the authority to withhold payments… 

Meanwhile, China keeps cranking out ships. So, this isn’t an academic exercise. We need to fix things. We need to stop talking and fix things, quickly, because there seems to be a war coming. The standard processes are not working. We need action now.