Overhaul 19e: Carrier Team One: Notes from an Unnatural Experiment
Introduction
Making Shipyards Weird (Temporarily)
The pace of improvement associated with Carrier Team One (CT1) was hard to sustain. With the exception of Toyota, few organizations sustain a culture of process inquiry and continuous improvement. It was especially exotic in shipyards (SYs), where the default was “do what the last guy did, but with more meetings.”
What made CT1 different? Over more than a decade as a participant, four things stood out for me.
First, CT1 applied systems thinking with discipline and enjoyed sustained senior-leader attention. Those two ingredients are essential for improving cross-organizational processes, and both are rare enough to qualify as endangered species.
Second, CT1 tried to make process improvement rewarding. I volunteered for as many Process Action Teams (PATs) as I could before the “Ralph Soule Rule” was instituted. Questioning the status quo was addictive.
Third, CT1 leaders armed their improvement teams with tools to overcome the inevitable resistance to change. The teams weren’t just told to “be innovative” and then left to fend for themselves against the bureaucracy’s immune system.
Fourth, the Executive Steering Committee (ESC) condensed its expectations into a compact anti-timidity metaphor: Dinner at the Umbergs.
Since 1997, much about the environment outside the shipyards has changed. That may be the most important reason the momentum for change initiated by CT1 couldn’t be sustained. Unnatural experiments seldom survive contact with the rest of the ecosystem.
On Sustaining Unnatural Acts
Systems Thinking vs. Blaming
Unnatural Acts are hard to sustain—Stein’s Law (paraphrased)
CT1 was “an unnatural act,” as one of its co-founders liked to say. Process improvement is an endless negotiation between two constraints: people capable of systems thinking and senior-leader attention spans. Optimizing a system is never finished because both the system and its environment are always changing. Meanwhile, senior-leader attention is consumed by crises, email, and the latest great idea from headquarters.
Start with systems thinking. If you don’t understand that what you are doing is part of a larger coordinated activity — with boundaries, inputs, outputs, and energy requirements — you have almost no leverage for improvement. If you can’t see the connections, if you don’t recognize that your output is someone else’s input, it’s easy to optimize your part of the process and quietly sabotage the overall system.
From my experience with CT1 and after, systems thinking is not widely practiced. There is a lot of uncertainty in organizations, but most people seem absolutely certain of two things: their small activity must function perfectly for them, and the people upstream are idiots (an attribution often extended to anyone in a leadership position). Their confidence might be shaken, slightly, if they discovered that the people downstream think they’re the idiots. Horrors.
The early CT1 leaders worked hard to block narrative fallacies from becoming the explanation for carrier overhaul problems. Narrative fallacies are the stories people tell themselves (and anyone who will listen) about what “those people” are doing wrong. The stories are simple (because they lack objective data), coherent (like flat-earth maps), and almost always wrong. The ESC decreed simple stories and blaming off-limits.
CT1 participants could only talk about processes, not the people in them. The beginning of process understanding is diagramming it: boundaries, inputs, outputs, and participants. One of Deming’s 14 Principles is that a key leader responsibility is to make sure everyone has an appreciation of the system. ESC leaders took this to heart at every meeting. Process-focused analysis consistently produced “aha” moments — and occasionally “you’ve got to be kidding me” astonishment — even from experienced carrier maintainers.
The second limiting factor for process improvement is senior-leader attention. Senior leaders, whose jobs tend to devolve into an endless sequence of crisis meetings, need “white space” to focus on the important rather than the merely loud. The early ESC leaders took the time to ask “Why?” and more time to listen to the answers. They invested their own time taking issues beyond their control to higher levels — time that could have been spent attending more status meetings (a favorite senior leader activity).
It’s possible that this level of engagement wasn’t sustainable. The SY managers and Type Commander leaders who learned to think differently moved on. ESC membership changed, key features of carrier maintenance changed, and the Navy policy environment changed. All of that was inevitable. Losing the habit of systems thinking was optional.
The Ralph Soule Rule
An Overachieving Lieutenant Commander
“How many teams are you on?”—CT1 ESC leader to yours truly
Warning: this section is shameless self-promotion, which made it fun to write and may be partially deserved.
Serving on CT1 Process Action Teams was empowering. I had suffered through many problems over two carrier overhauls. Processes didn’t work well, asking why wasn’t encouraged (too much work to do), and getting anything done required an impressive array of workarounds. Most involved hiring additional people at Type Commander expense and claiming that it was a “lesson learned.”
Before the first CT1 meeting in February 1997, I spent hours with teams identifying shipyard process problems, writing point papers, and recommending changes. It was exciting. SY customers actually cared what we thought about improving overhaul performance. I was going to tell them, by golly, whether they wanted to hear it or not.
After the February meeting, my superiors let me participate on as many PATs as I wanted. I think they didn’t know what I was doing! Most of the teams made bold recommendations for change, but some of the civilian members were worried about presenting them in an open forum at the follow-up meeting. Presenting radical ideas to senior leaders is risky, especially if you plan to keep your job.
Naval officers enjoy special license in SYs to shake things up ask a lot of questions. This freed me to volunteer to present all the crazy stuff from four PATs. Fortune favors the brave. In this case, it favored the naïve.
After my fourth briefing at the next CT1 meeting, the ESC Chairman asked, “How many teams are you on?” That is not a question you want asked if your chain of command can hear it.
Several ESC members wondered aloud whether it was wise to have one person on so many teams. “What if he gets hit by a bus?” This concern was codified as the Ralph Soule Rule: no one could be a member of more than three PATs. Why three was better than four or worse than two was never explained. It did, however, confirm an important principle: in process improvement, risk management starts with controlling enthusiastic Lieutenant Commanders.
Overcoming Obstacles
Beating the “It’ll Never Work” Chorus
The CT1 PATs faced strong resistance to change. Asking people to do things differently is unpopular. No one could have foreseen that! The ESC provided three particularly useful tools to overcome resistance: systems thinking, narrow scope, and pilot projects.
In this context, systems thinking meant looking at elements, relationships, and purposes in complex environments and how they interact to generate patterns of behavior. CT1 improvement efforts proceeded by identifying gaps between actual and desired outcomes, proposing the shortest path to close them, eliminating stupid stuff (usually based on faulty assumptions, not stupid people), documenting the revised process (with a diagram), rewriting applicable Navy instructions, and training for the new process — the most neglected step in “change initiatives.” No wonder people dislike change.
The ESC’s focus was on identifying where value was created and removing everything else. For the PATs, it was like The Emperor’s New Clothes. The ESC gave them permission to point out the nakedness in carrier maintenance. Like a visit to a nudist colony for people over the age of 50, sometimes it wasn’t pretty, but once you saw it, you couldn’t unsee it.
The second tool was keeping a narrow focus. When something in ship maintenance looks simple (“Why do we ask the crew to agree to protect piping from freezing?”), people can get excited. Well, maybe not about freeze protection, but you understand the idea. Ship maintenance excitement (yes, it’s a thing) produces reactions like, “This is great. Why doesn’t everyone do [fundamental maintenance process] this way? Everyone in the Navy has to do this.”
The narrow focus of CT1 process improvement turned out to be one of the ESC’s best ideas. They recognized that Navy ship types and their maintenance processes were too diverse and complex for everyone to do things the same way (that hasn’t changed). They also recognized that they didn’t have the power or energy to analyze and change everything, everywhere, all at once. For a chosen few, reality is an effective constraint.
By keeping the focus narrow, the ESC authorized PAT leaders to use the “not at this time” defense. Whenever a third party insisted that the PAT’s ideas be replicated beyond carriers, team leaders could say, “You may be right, but we’re starting with aircraft carriers. We’ll get back to you if it works the way we expect.” Secret translation: if this goes badly, at least we’ll know where the crater is.
A third obstacle to change was the “It’ll Never Work” chorus. This is a variation on, “…we don’t have a definitive solution to our long-term problem. What we do know is we don’t like yours.” People don’t usually state it so clearly, unless they are the Secretary of the Treasury.*
* U.S. House of Representatives, Committee on the Budget. 2012. “The President’s Fiscal Year 2013 Budget: Revenue and Economic Policy Proposals.” Hearing, 112th Cong., 2nd sess., February 16.
The ESC provided high cover against doom forecasts. PAT leaders were authorized to reply, “This isn’t a permanent change. We are proposing a trial, will test it on one ship’s availability, and the ESC will evaluate the result. They’ll let you know how it goes.” The pilot improvements were narrowly focused and identified specific data for evaluation. That made opposition much harder to sustain — it’s difficult to argue against focused change by waving your arms. Especially when the other guy has a plan.
Dinner at the Umbergs
Killing the “Do What the Last Guy Did” Protection Mechanism
Process improvement requires boldness by the truckload. Boldness to call the emperor naked. Even more boldness not to look away. Boldness to propose new processes. Boldness to deal with “just say no” bosses who don’t want to accept responsibility for doing something new (something could go wrong). Many people would prefer to do things they know don’t work than accept responsibility for doing something new that *might* work.
Bureaucracies managing complex work can diffuse responsibility. If eight people have to sign off on a decision, who was in charge? Some, relying on mind-reading rather than evidence, claim it’s all deliberate—just a way to dodge responsibility. I believe much of it is a consequence of the normal work and specialization inherent in bureaucracies. Leaders use hierarchy, control, and groups of technical experts to produce consistent (if boring and inefficient) outcomes. Sometimes that is good enough. It falls short when there is pressure to reduce cost and deliver outputs faster.
Boldness, the courage to try new things, and learning from mistakes are not widely valued in bureaucracies. That has less to do with organizational design than with the humans in it. Even when managers say they want people to think differently, they seldom act like it: criticizing new ideas, avoiding risk, and punishing people when things go wrong. This is called “accountability” and works great unless you want to learn what went wrong and how to do things differently in the future.
Conservatism and accountability in organizations aren’t sins. When the risk of meltdown is high, they are virtues, even when they are expensive. SY repair processes are conservative approaches to producing reliable maintenance outcomes for high-threat environments. Doing what the last person did won’t lower cost and might take longer, but it is personally safer in SYs. Fortune may favor the brave, but job security favors the cautious.
SY conservatism was a major challenge for getting CT1 members to think differently and act bolder. For reasons beyond the scope of this blog, SY culture had become so conservative by the 1990s that it was hard to get things done. Costs weren’t going down either. Depot maintenance was very expensive. Before CT1, it was safer and more logical for SY managers to hunker down and wait to be told what to do. Creativity was something you did on weekends.
Dinner at the Umbergs was an ESC metaphor for encouraging boldness. It was based on a three-panel Blondie cartoon strip by Chic Young. Young lampooned the “do what the last guy did” mentality. The strip showed a couple planning a dinner party. They wanted it to be just like the one they attended at the Umbergs: same menu, same guests, and even at the Umbergs’ house!
ESC members used the Young comic as a compact way to prod PATs and project teams to be bold. They didn’t want an availability plan to be “Dinner at the Umbergs” — an imitation of whatever the last project did, regardless of whether it worked, failed, or merely produced the same old overruns on schedule and cost.
Conclusion
Like old soldiers, unnatural acts just fade away
We got lucky with CT1: the right people, at the right time, aligned with interests and finances to fight transparently stupid processes — stupid in ways that became obvious once the process diagrams were on the wall. The ESC leaders were superbly prepared Navy leaders who invested in long-term relationships with the only shipyards capable of repairing their ships. They were willing to spend their political capital on better processes instead of more briefings.
As I’ve noted in previous posts, the causes of bad overhaul outcomes haven’t changed much since 1997. What has changed are the workforce (numbers and experience), work package sizes, political constraints, and budgets. The Navy warship maintenance system has changed significantly since the early years of CT1. The incentives to cling to “Dinner at the Umbergs” practices have not.
CT1 may have progressed as far as it could while remaining external to NAVSEA, the naval shipyards, and the exigencies of fleet operations. Managing and improving availability outcomes are collective activities that require sustained senior-leader time and attention to stay aligned. Alignment in the Navy can be as scarce as real coordination at a demolition derby unless the Chief of Naval Operations says, “Do this now!” That’s why CT1 felt like an “unnatural act.” The CNO doesn’t always get what they want either.
My next post will be about the SSN 688-class Baseline Project Management Plan. It was a different, and very effective, way to approach submarine depot maintenance improvement. Like CT1, it likely can’t be repeated because the world is different, but it remains an instructive example of what happens when systems thinking is allowed out of the broom closet and given permission to rearrange the furniture. Eventually, someone shoves it back in the closet, right next to last year’s transformation initiative.
