Something extraordinary is happening on the Texas coast. The world’s largest rocket, quite literally built in a tent, is preparing for its first flight. This rocket, built by SpaceX and called Starship, is about twice as large as the Saturn V, and will aim to be fully and rapidly reusable…the “holy grail” of rocketry.
Elon Musk, CEO of SpaceX, is known for his unique approach to problem-solving. His approach embodies his years of experience across multiple industries, each one he disrupted and overturned. Here is a comprehensive look at Musk’s insight’s, insights that you may apply in your own life.
Step 1) Make the Requirements Less Dumb
Musk recently gave an interview where he outlined five steps that each person working on Starship must take daily. The first step is to make the requirements “less dumb.” This is another way of saying, “asking the right question” or “defining the problem” correctly.
Musk has often mused that when approaching the design of reusable spacecraft, it took years to properly define the problem, but that once the problem was well defined, the solution followed.
Those solutions, according to Musk, should arise from First Principles and not though analogy. That is, when designing a car, don’t look to other cars and improve upon them, approach the problem as if you are building the first car ever conceived.
Musk famously utilized First Principles’ reasoning with his first rocket, the Falcon 1. When told that a private company could never afford to design and build their own rocket, Musk grabbed a pen and paper and listed the raw materials used in rocketry alongside their prices, then added them together.
He demonstrated that, in fact, rockets weren’t very expensive at all. Rockets designed and built traditionally were expensive, but there was no fundamental reason they had to be.
Step 2) Delete the Part or Process
Musk has learned a lot from the assembly lines of Tesla, and these lessons have been applied to Starship. The Starship design has been getting simpler by the month, as parts and processes are deleted. This all makes for a cheaper and safer rocket.
But it’s not always obvious which parts can be done away with, partly because of organizational limitations. Musk revealed how he overcomes this in 2019 when he noted that an organization’s product often resembled the organizational structure that designed it. Musk may not have known it, but what he was describing is known as Conway’s Law.
The organizational structure creates points of weaker communication. To use a Tesla example, Musk has frequently lamented that he spent a great deal of time attempting to automate a machine used to place a felt cover over the battery pack of the car before it was raised up into the body assembly.
After repeated failures to get this machine working, he asked the battery team what the purpose of the felt cover was, they informed him that it was to reduce sound and vibration. The auto body team, however, when asked the same question, replied that it was for fire safety.
It turned out that the part was not needed at all and could be deleted entirely. The part existence was a product of poor communication between disparate teams.
Lesson learned, Musk overcomes this on Starship by having a flat organizational structure with each person serving as their own chief engineer while minimizing teams and organizational divisions that create communication bottlenecks.
Step 3) Optimize
Some parts and processes are not negotiable. This is where traditional engineering comes into play. Processes can be improved and designs can be optimized through modelling and testing. This process of optimization, however, is closely linked to Step 4, as we will see.
Step 4) Accelerate Cycle Time
In recent years, Musk has talked a great deal about the interplay between production and design. For all of its complexity, it is relatively easy to design and build a single car or rocket, producing them on scale, however, is at least 1000 times harder according to Musk.
Why bother mass-producing rockets at all then? Well, there are a number of answers to this question, but it turns out that design optimization is inextricably linked with production cycle time.
Musk notes that technological progress follows the below formula:
Technological Progress = Iterations x Progress Between Iterations
In other words, the more quickly one can iterate and test, the more quickly flaws and optimization opportunities can be revealed. That is why Musk is comfortable with rocket explosions and failures…they reveal important data that help improve the end product faster.
This is akin like the parable of the pottery class. When a teacher divided his pottery class into two groups, grading Group 1 on quality and Group 2 on quantity, the second group, despite not even trying, would produce a better quality piece than the first.
Why? Because the second group, in producing in quantity, had more iterations to improve quality. It’s the same reason that a Camry is more reliable than a Ferrari; mass production creates a stronger feedback loop for design optimization.
Step 5) Automate
This last step is obvious. Once you have developed the optimal solution and refined it such that you can achieve excellent cycle times, the next logical step is to automate. Automation can reduce the cost both in terms of expense and human labor, freeing up resources for other areas where it is now needed.
The above steps aren’t limited to rocket production. From writing, to your job, to government, to software design, the basic principles identified by Musk are widely applicable. Musk has been able to leverage his experience in multiple industries and has been willing to share his experience for the betterment of our personal lives and the world as a whole. We best heed the lessons.