As a lover of all things space, I was super excited to watch a series of YouTube videos published by Tim Dodd (Everyday Astronaut) where he was given a tour of SpaceX’s StarBase facility in Boca Chica by none other than Elon Musk. While I got the expected doses of geeking out over rocket science and an appreciation for the sheer enormity of Starship, I also got something unexpected: a lesson on process building.
Throughout much of the series Tim and Elon spend time discussing production and process of the various components of Starship, the Falcon rockets, and even Tesla vehicles. Elon is definitely one to tout the importance and challenge of production, stating earlier this year in a tweet:
“Design is easy, production is hard”
https://twitter.com/elonmusk/status/1411971491319586820?lang=en
In fact, last year, in a response to a tweet from Tim, Elon states:
Creating the production system is >1000% harder than building one rocket. This is the truly hard thing.
https://twitter.com/elonmusk/status/1269434513085485058?lang=en
Production, in this case, is simply the process by which the product is manufactured. Creating that process is the challenge.
During their tour, Tim asks Elon if the Starship grid fins are going to fold up. Elon’s answer is a masterclass in process building. He describes a 5 step process (outlined in detail below) that can be used in any industry when developing or evaluating a process.
Step 1: Question Requirements
Or as Elon puts it: “Make Requirements less dumb. It doesn’t matter who gave them to you. It’s particularly dangerous if a smart person gave them to you because you might not question them.”
Here Elon highlights the importance of questioning. “Everyone is going to be wrong sometimes,” he says, so it must be not only acceptable, but expected that all team members are prepared to question requirements.
Elon adds, “whatever requirement or constraint you have, it must come with a name, not a department because you can’t ask the department, you have to ask a person. The person putting forward that requirement or constraint must take responsibility for that requirement or constraint.” This detail is about having accountability. Having a person taking ownership of a requirements means there is a clear place to go when questioning that requirement. Without this clarity of responsibility Elon says “you can have a requirement that came from an intern two years ago that is no longer with the company but it came from the air load department, even though no one currently in air load department agrees with the requirement.” The person responsible for the requirement must also be held accountable to the requirement when it is being questioned.
Step 2: Delete
“Try very had to delete the part or process,” Elon says. He explains that we naturally have a bias toward “in case we need it” which is often to the detriment of the process. Unnecessary steps or parts can add unneeded time, cost or complexity. Elon encourages his teams to “fight against that, don’t hedge bets.”
As a measure, Elon explains “if you’re not occasionally adding things back in you’re not deleting enough.” You need to remove parts and steps to not only eliminate what is unnecessary, but also to confirm the necessity of those that will remain.
Step 3: Optimize
Optimization is often one of the first process improvements organizations attempt. However, Elon cautions against optimizing before first questioning and deleting. He says, “One of the most common errors of a smart engineer is to optimize a thing that should not exist.”
We often try to make a step or part better, faster, cheaper, easier, without ever questioning do we even need this step or part? Elon explains we have been conditioned to behave this way because “everyone has been trained, in high school and college that you gotta answer the question, convergent logic. So you can’t tell a professor, your question is dumb. You will get a bad grade, you have to answer the question. So everyone is basically, without knowing it, they’ve got like mental straight jackets on that is they’ll work on optimizing the thing that should simply not exist.” It is important to simplify a process, but only after first ensuring you’ve tackled the requirements and also removed any unnecessary components.
Step 4: Accelerate
Next Elon says, “Accelerate cycle time. You’re moving too slowly, go faster.” This is optimizing specifically for speed. How can the process be improved so that we get the quality results quicker?
However, be careful not to go faster until you’ve first completed steps 1-3. As Elon says, “If you’re digging your grave, don’t dig it faster, stop digging your grave.”
Step 5: Automate
The final step is to automate. Take a manual process and develop ways to make it happen on its own. This could mean reducing interaction and interventions in the process or leveraging technology (software or hardware, AI, robotics) to perform steps.
In Process Testing vs. End of Line Testing
In addition to the 5 steps for process improvement, Elon notes that processes are often hindered by “too much in-process testing” to isolate a mistake. He says “ You don’t know where things are breaking so you’ll test work in process at various steps to isolate where the mistake is occurring.” This is a common diagnostic practice to help determine which steps might require improvements. The danger is in the delays that are created if these checks are left in place long after they are no longer needed. Elon explains, “A very common mistake is to not remove in process testing after you’ve diagnosed where the problems are. If things pass at end of line testing, you can cut out in-process testing.”
The Order Matters
Elon explains that he has made the mistake of going backwards on all of these steps. He shares an example involving fiberglass mats for the Tesla Model 3 that were installed between the floor pan and the battery. The installation of these was choking the Model 3 production line and Elon was spending day and night on that part of the line trying to make a correction. First he tried automating – making the robot better. Then he tried accelerating: making the robot move faster, have a shorter path, increase the torque. Then he tried optimizing, and instead of spackling glue for the mat, to just put little dabs of glue to hold it in place until the battery pack was bolted into the car. However, none of this created a significant enough improvement to address the delay in the production line.
Automating was a mistake.
Accelerating was a mistake.
Optimizing was a mistake.
Finally Elon asked, “What are these mats for?” When asking the battery safety team they told him the mats were for noise and vibration. When asking the noise vibration analysis team they stated they were for fire safety. Each team thought the other required the mat, but neither department thought it was necessary for their own purposes. So finally, after questioning the requirement, they tried the car with fiberglass mats and without and couldn’t tell a difference as far as noise and vibration were concerned so Elon says they “just deleted them and bypassed this whole two million dollar robot cell as a complete pile of nonsense.” Question the requirements, delete the part, move on to the next problem.
Since hearing Elon explain his methodology for approaching process improvement I’ve already noticed processes within my team that we needlessly have been trying to optimize that should more than likely be deleted. I find we have also not done enough to question requirements in a process. Needless requirements eliminate viable solutions to process problems prematurely. By ditching unnecessary requirements, we define a cleaner problem and therefore a cleaner solution.
tl;dr
So in summary, if you want to build or improve a process you must:
Question Requirements
Delete parts/steps
Optimize
Accelerate
Automate
You can check out Elon discussing these 5 steps on Youtube:
https://youtu.be/t705r8ICkRw?t=806
If you’re interested in space, rockets, engineering, SpaceX, or Elon Musk I encourage you to watch all 3 episodes of the Starbase Tour series from Tim Dodd’s EverydayAstronaut channel:
Part 1 – https://www.youtube.com/watch?v=t705r8ICkRw