3 ms·
> they are pretty much close to the promised rate of 5K cars/week They might hit that goal, but I'm very concerned about how they're getting there. Toyota was
by wpietri 8y ago
> they are pretty much close to the promised rate of 5K cars/week
They might hit that goal, but I'm very concerned about how they're getting there.
Toyota was the company that really figured out how to make good cars economically. They came out of Japan's post-war economic devastation to kick the asses of US carmakers. Their thinking was so alien to traditional approaches that Detroit spent decades trying to catch on and catch up. That depite Toyota doing their best to help their competitors get it. [1]
A pillar of Toyota's approach is patient, gradual, relentless improvement. There are a lot of signs that Tesla has never really warmed to that. Musk likes drama, action. But he's been overpromising and underdelivering for a while now. That he's sleeping at the factory might sound impressive, but to me it's a sign that he hasn't built an organization that can handle this.
Even if they hit the arbitrary 5K goal, I'm concerned they'll will reach it not by building a great, reliable process, but through the sort of rushing and corner-cutting that leaves them in bad shape for the future. That's very common in the tech world, and we pay for it with high tech debt and low long-term productivity. Perhaps he's doing the same thing here.
[1] e.g., https://en.wikipedia.org/wiki/NUMMI https://en.wikipedia.org/wiki/NUMMI
- davedx 8y ago> A pillar of Toyota's approach is patient, gradual, relentless improvement. There are a lot of signs that Tesla has never really warmed to that. I guess it depends how you interpret the news and what Musk has said in shareholder meetings. At the last one he explicitly said that they continuously improve the process at Tesla.
- wpietri 8y agoI'm sure he said that. But the US car industry has a long history of boldly declared but ultimately failed attempts to adopt Toyota's process in a way sufficiently deep to get Toyota's results. Rother's book "Toyota Kata" is a good dive into why: they adopted some of the rituals, but never made the deeper cultural and philosophical changes needed to get most of the value.
- windexh8er 8y agoToyota embraced Deming/Kaizen, it wasn't their idea. Beyond that Tesla is a more extreme version of those ideas. Toyota doesn't need to, but it's likely they also wouldn't, build an entirely new line to meet a goal. That would be out of the focal range of their 'softer' implementation of continual improvements. Musk on the other hand is not averse to continual improvements that are on the extreme end of the spectrum. This is very evident in how they've built the car. For example when Consumer Reports identified a braking issue with the Model 3 Tesla had an OTA software fix in days. What other manufacture would do that, let alone could?
- jcranmer 8y ago> For example when Consumer Reports identified a braking issue with the Model 3 Tesla had an OTA software fix in days. What other manufacture would do that, let alone could? Why should that be a good thing? Consider what this means: * Safety-critical systems are not air-gapped. (Granted, Tesla is far from the only offender here). * The fact that the fix was turned around in such a short time would seem to suggest that it was relatively quick to diagnose and debug. Why wasn't it caught before release? * On another note, it also calls into question how rigorous their update fix testing is. I've seen plenty of "this must go out fixes" that broke a lot of stuff by collateral when people felt it was too important to get out quickly to do the full testing.
- mijamo 8y ago- As you said, not the only offender here, I would even say air gapped would be the exception - This is wrong, the fact that is was quick to fix does not mean it was easy to find, or even easy to fix. Many softwares have suffered critical problems for ages without being detected, and yet when they are detected they are fixed in a day? - Sure, you can wonder, but this is purely speculative as we have no evidence of any problem of the sort here.
- planteen 8y ago> the fact that is was quick to fix does not mean it was easy to find The issue was easy to find. The root cause may not have been. Vehicle stopping distance is a standard metric. Simple black box system level testing at Tesla should have caught this. Why was nothing done then or why was it a test escape?