People are beginning to notice that Tesla has problems, and that many of them are symptomatic of the dot-com ethos of its founder, Elon Musk.
The most classic example is that of their manufacturing problems, where they are seeing both quality and productions problems, and now an analysis of Tesla’s production plans show that both are a function of its single minded insistence on eliminating the human component:
The robots are killing Tesla.
In a rare win for humans over robots in the battle for labor efficiency, Wall Street analysts have laid down a compelling argument that over-automation is to blame for problems at the billionaire Elon Musk’s electric-car company.
That is to say, the very innovation and competitive advantage that Musk says he’s bringing to the car industry — his nearly fully automated plant in Fremont, California — is the reason Tesla is unable to scale quickly.
According to the Bernstein analysts Max Warburton and Toni Sacconaghi, it’s the robots that can’t pump out Tesla’s highly anticipated Model 3s fast enough. The whole process is too ambitious, risky, and complicated.
From Bernstein (emphasis ours):
“Tesla has tried to hyper-automate final assembly. We believe Tesla has been too ambitious with automation on the Model 3 line. Few have seen it (the plant is off-limits at present), but we know this: Tesla has spent c.2x what a traditional OEM spends per unit on capacity.
“It has ordered huge numbers of Kuka robots. It has not only automated stamping, paint and welding (as most other OEMs do) — it has also tried to automate final assembly (putting parts into the car). It talks of two-level final lines with robots automating parts sequencing. This is where Tesla seems to be facing problems (as well as in welding & battery pack assembly).”
Bernstein adds that the world’s best carmakers, the Japanese, try to limit automation because it “is expensive and is statistically inversely correlated to quality.” Their approach is to get the process right first, then bring in the robots — the opposite of Musk’s.
It’s not a problem that Tesla, a highly indebted company, can afford forever.
So in Musk’s attempt to bring on the robot uprising that will revolutionize how we make cars, he’s burned cash and baked in his own mistakes. If you think about it that way, we are just beginning to understand how much this will cost him.
The dotcom ethos is ship it then fix it, and that if you have a problem, you can always throw a few more servers and few more programmers at this.
So, Tesla has been trying to do something that the auto industry has been trying, and failing, to do for years, and its solution is to throw more hardware at it without addressing the underlying problem.
Very dotcom, and it’s why they continue down their dysfunctional path, because this is how it is done in IT. (Also why every few months you hear of a data breach)
Here is a money quote from the report, courtesy of Naked Capitalism:
[A]utomation is expensive — and usually proves far less effective, highly inflexible, and creates quality problems further down the line.
- In welding, mistakes and inconsistences go unrecognized — but the machine powers on and builds cars with the wrong geometry or bad spot welds in key locations. These are only found later — when for instance the windscreen is inserted, or a door re-attached. Have you wondered why Tesla doors don’t align, or hoods don’t fit, or windscreens are prone to cracking? Now you have your answer.
- In final assembly, robots can apply torque consistently — but they don’t detect and account for threads that aren’t straight, bolts that don’t quite fit, fasteners that don’t align, or seals that have a defect. Humans are really good at this. Have you wondered why Teslas have wind noise problems, squeaks and rattles, and bits of trim that fall off? Now you have your answer.
This is a state of affairs that has the New York Times wondering about its continued viability, and Bloomberg has noted that Tesla bonds are tanking.
Software can facilitate some aspects of operating in the real world, but when the primary activity is operating in the real world, it cannot change reality.
Overselling and under-delivering is a feature, not a bug.