Table of Contents
This won’t be the first time that I am writing about Elon Musk. There is a good reason for this. Like many other technology nerds, I am infatuated with Elon’s ability to approach some of the worlds largest problems systematically and methodically. In a recent interview Elon spoke to Tim Dodd about the 5 Manufacturing Principles that govern the way that SpaceX approaches the development of the Starship.
Before I outline the 5 principles, I think it’s important to provide some context. At the time of writing this, SpaceX has just recently been awarded a massive contract with NASA to provide the HLS (Human Landing System). SpaceX will be responsible, along with NASA, for getting people back to the moon for the first time since the Apollo era. SpaceX has gotten a lot of flack from the likes of Amazon for it’s seemingly complex systems. Here is where SpaceX is different to any other space company that has come before. SpaceX isn’t just trying to create one single viable rocket like NASA’s SLS (Space Launch System). In other words, they’re not just doing enough, SpaceX wants to make the rapid manufacturing of Cheap Reusable rockets a possibility.
Elon’s approach to SpaceX’s manufacturing process is no doubt thanks to years of experience that he has gained solving issues on the Tesla manufacturing lines.
I’m not in manufacturing, why should I care?
The lessons that we can learn from this are valuable. While this is specific to the scalability of a manufacturing process, we can apply these rules to our own lives. Especially, in the work place. The reality is that the systems we create determine the behaviour of those who are at mercy of the system.
The manufacturing problem
One of the problems with scaling a manufacturing process, is that you’ll often have multiple teams developing different parts in isolation. Elon has a great anecdote that demonstrates the issue with development in isolation: There was a process in the Tesla Manufacturing process that required a robot to glue a thermal mat to the Tesla’s battery. This process was a choking point in the entire manufacturing process of the Tesla. Elon tried various different approaches to try alleviate the choke point. He tried to speed the process up by making the robot go faster. He made the robot apply less glue. Eventually he asked the department in charge of fire safety the reason for the thermal mat. Their response was that it was for noise and vibrations reduction. When Elon spoke to the team in charge of noise and vibrations, they said it was for fire safety. When they removed the mat there was no notable difference in the noise an vibrations and it didn’t pose any threat to fire safety.
Step 1 – Make your requirements less dumb, your requirements are definitely dumb.
- Always question requirements, especially if they came from a smart person.
- Requirements should come with a name, not a department. This way you can talk to a person directly about why that requirement exists.
Step 2 – Try very hard to delete the step or process.
- “If you’re not occasionally adding stuff back, then you’re not deleting enough” – Elon Musk.
Step 3 – Simplify or Optimize
It’s the most common error of a smart engineer, is to optimize something that shouldn’t exist. The reason for this that convergent logic is programmed into us from a young age. Convergent logic is when you are given a whole set of facts and asked to solve a problem. The problem with this is that you can’t use “out of the box thinking” or lateral thinking. Instead of asking “what is the answer to this question?”, ask: “should we be answering this question?”
There is a maxim in SpaceX and Tesla engineering circles: “No Part, is the best part. Elon talks about how smart engineers will try optimize something that doesn’t need to exist in the first place. A good example of this would be the grid fins on Super Heavy Boosters. At first they were depicted to fold against the rocket when not in use, but in reality this would create a layer of unneeded complexity. So, when the first grid fins were attached they didn’t fold against the rocket.
Step 4 – Accelerate Cycle Time
- Don’t go faster until you’ve mastered the previous 3 things.
- Like in the example of the Tesla thermal mat, trying to speed things up isn’t going to help much if the process can be eliminated or simplified.
Step 5 – Automate
- This is self explanatory. Once you’ve done the the above 4 steps, try to make them work with as little human intervention as possible.
I hope you found this helpful. I am going to turn this into a slightly more in depth blog post at a later stage. Let me know if there was anything that you found interesting or surprising. Have an awesome weekend!