Internal documents reveal the grueling way Tesla hit its 5,000 Model 3 target

https://www.thisisinsider.com/tesla-hit-model-3-target-by-reworking-thousands-of-cars-2018-8

by plasticchris • 8 years ago

191 171 8 years ago

kenjackson• 8 years ago

If this impacts the total time to make each car, wouldn't that impact your production volume? It seems weird that something that should make you slower is what allows you to hit your target? Unless they're saying the target doesn't take this into account -- which would just seem stupid.

mikeash• 8 years ago

Rework can happen in parallel with production, so it won’t necessarily affect your rate.

kadenshep• 8 years ago

It definitely affects labor productivity.

ben1040• 8 years ago

I'm wondering if what they're saying is that there was a large enough quantity of to-be-reworked assemblies from a prior week that reworking them and calling them done gave them enough "completed" vehicles to put them over the target for a later week.

dgritsko4• 8 years ago

It's not that it affected the time to make each car, it's that a large majority of the cars needed to have some issue addressed after they were manufactured (86% according to the article).

michaelt• 8 years ago

If your production line has 10 machines that break the car windscreen, the long-term most efficient thing would be to fix those 10 machines.

But that'll take time, and in the short term putting a guy at the end of the line replacing all the windscreens gets cars out the door today.

setr• 8 years ago

I think what it means is that they gave up on correcting the automation process in favor of just pumping out cars and fixing them up manually. The former having longer-term benefits, but stalling car production until its complete. In other words, they just threw more bodies at the problem to get it out asap

wongarsu4• 8 years ago

I think the title "Tesla hit Model 3 target by reworking thousands of cars" should be read as "Tesla hit Model 3 target by not only producing the cars, but also reworking them in time".

It's confusing, I prefer the article's headline "Internal documents reveal the grueling way Tesla hit its 5,000 Model 3 target" (the submission title is from the URL)

gsnedders• 8 years ago

Classic case of throughput (i.e., how many cars are completed per month) v. latency (i.e., how long it takes one car from start to end) — the targets are almost always about throughput and not latency.

xkjkls10• 8 years ago

Tesla was always accused of some chicanery when they claimed to hit a 5000/week of "factory gated" cars -- a term they had never used before. Also, last quarter, they had more than 10,000 cars "in transit" between the factory and delivery, which they didn't fully explain how much re-work might be needed on before a customer can accept delivery. It's part of a pattern of a company being misleading with its metrics.

GuB-42• 8 years ago

So what? They talk about it like it is a bad thing.

Their cars work, they met their target, who cares about the details of the manufacturing process? Yes, there is room for improvement, there is always room for improvement, I let the guys at Tesla deal with their priorities.

It is a nice insight on how things are done at Tesla but that's about it.

deelowe• 8 years ago

It's terrible from a supply chain and logistics standpoint.

dgritsko4• 8 years ago

Manufacturing rate is one measure of success. First Pass Yield (FPY - % of cars that don't require rework) is another. It seems like Tesla is sacrificing the latter in service of the former, or at least they were towards the end of Q2. I want to see Tesla succeed, but it seems like they really need to focus on increasing FPY in order to improve their long-term odds of success.

wongarsu4• 8 years ago

What makes FPY an important factor seperate from manufacturing rate? (if we assume a car is manufactured once it is ready to ship)

Of course it impacts your manufacturing rate, since anything not passing on first pass needs rework and possibly additional parts, making logistics more difficult etc. But all of that just influences manufacturing rate and cost per vehicle.

sandworm1014• 8 years ago

>> they really need to focus on increasing FPY in order to improve their long-term odds of success.

They really need to start being a car company rather than an investment factory. Car companies, particularly the BMWs and Hondas Tesla wants to compete with, care more about their reputation than factory output. I cannot imagine a BMW exec ever sacrificing unit completeness in the name of productivity. For Honda, an unfinished car simply isn't a car worth counting. Tesla has a long way to go before winning me over. (also Toyota)

stinos• 8 years ago

First Pass Yield (FPY - % of cars that don't require rework) is another

I get that, but isn't it expected that if you create a new car factory, sort of from scratch, your FPY won't be exceptionally high? Unless you really know what you're doing due to previous experience in car manufacturing?

cm2187• 8 years ago

For a cash strapped company that critically depends on delivering large enough volumes to survive the year, I can see why many investors would be interested in the nature of the headline numbers announced publicly. Otherwise I agree that it normally takes years to adjust the production of a new product, particularly for a young company, and it is unreasonable to expect high yields on year 1.

foobiekr2• 8 years ago

High rework rates are indicative of overall quality issues. Issues caught and flagged for rework are the obvious ones, not all of the defects.

A high rework rate is a huge red flag for at-scale manufacturing and says that something is wrong.

It seems like this should be obvious. How healthy is a software engineering environment where 80% of the commits fail at the tail end of the development pipeline and require fixes? I’ve never seen an org that bad.

jayd162• 8 years ago

High rework rates are indicative of overall quality issues

Maybe, maybe not. We can look at the actual quality of the new cars instead of trying to guess.

wemdyjreichert• 8 years ago

Quality issues? Hmm, possibly related to bumpers falling off in rain.

compcoffee2• 8 years ago

So what? They talk about it like it is a bad thing.

How do you expect them to be "world leaders" in manufacturing technology when they have to rework so many cars? Remember, the narrative is that no one can do it as well as they can. Huge margins!

who cares about the details of the manufacturing process?

Yeah, who cars about the manufacturing process when your whole business is manufacturing...

LoSboccacc4• 8 years ago

who cares about the details of the manufacturing process?

buyers worried for assembly quality and overall reliability?

for example a misplaced or improperly installed mounting brakets make panels noisy over time, panel misalignment will affect resale value, badly tightened glasses will opacize etc.

people aren't able to see the car they're going to get beforehand, at most they can refuse upon delivery, se yes knowing assembly line issues is kinda important for a car, not everyone is a valley software engineer pulling 150k+

edit: wow musketeers out in masses I see

RobLach• 8 years ago

”who cares?”

Investors

titanomachy• 8 years ago

37 minutes of rework per car doesn't sound like very much. It would be interesting to know what fraction of the total time that is.

dgritsko4• 8 years ago

37 minutes * 4300 cars = 2,651 hours of work.

strig2• 8 years ago

So basically one full time employee for a year.

jakelarkin• 8 years ago

2,651 hours/week * $30/hour = ~$80k/week. on a volume of cars that brings in $200m/week. so what? its a minor inefficiency.

kingbirdy• 8 years ago

Per [0] it seems like line workers make ~$20/hr, so that's over $50,000 per week to fix the cars, not including the cost of parts or associated delays. It's only about $12 per car, but for Tesla I imagine every dollar counts.

[0] https://www.indeed.com/cmp/Tesla/salaries?location=US%2FCA%2...

user5994461• 8 years ago

It's a bit more than the output from a single person in a year. You just have to hire 2 more guys and you're good.

MBCook4• 8 years ago

No, but if 85% of cars need it that adds up fast.

nsxwolf• 8 years ago

Is this expected to be how things are going to go for the foreseeable future? If the situation is improved quickly, I don't see how this matters at all.

gamblor9568• 8 years ago

As noted in other comments, other major carmakers can complete a car in about 1-3 minutes, not including customization tasks like painting.

So 37 minutes of rework per vehicle means that they're less than 5% as efficient as their competitors, despite the massive investments they've made their factory line. It's indicative of serious production issues and will be a major roadblock to scaling future production, since it means that if they don't solve the production issues leading to the rework, then they would need to spend 20x what their competitors to achieve the same volume of production.

EDIT: The 3 minutes is time-between cars. Each car takes 17-18 hours to get through the entire line. So 37 additional minutes is still pretty bad, but it's not 20x as inefficient on a time-basis. However, given the number of workers at the Tesla factory, they are still 20x as inefficient on a labor basis.

freeone3000• 8 years ago

That's pipelined time, not per-car time. Just because a car is produced every 3 minutes does not mean it takes 3 minutes to produce - that would be incredible.

jessewmc• 8 years ago

Are you sure? Not a production expert, but it sounds to me like this is a confusion of 'complete a car every 1-3 minutes' and 'complete a car in 1-3 minutes' They are two very different things, and I'm quite sure that no one is completing entire cars in 1-3 minutes.

rurban2• 8 years ago

It's not bad at all, because it is only needed once until the bug is fixed. It's only a wrong manual task, so fixing this task sounds even trivial.

Const-me• 8 years ago

Exactly.

Tesla says they pay $25/hour at Gigafactory, glassdoor says they’re paying $17-$22 hours to manufacturing workers.

I’m not into manufacturing, but extra $10-15, for a car with starting price $35k, looks normal.

rurban2• 8 years ago

It's also fixing a manual part, not an already automated part. Remember, Tesla has a much higher automation factor than everyone else, esp. to get the quality up and costs down in the end.

Fixing just this failing manual part is not the drama the competitors are declaring. Tesla also needed this artificial 5.000 number to reach their mandatory milestone.

wongarsu4• 8 years ago

14% of the vehicles made didn't need rework [...]
The most common reason Model 3 cars require rework is a "failed manual task," followed by cosmetic issues

Given how entirely ad-hoc their entire production line still is, and how the car doesn't seem to be designed for easy assembly at all, this is unsurprising. But at least they don't seem to skimp on quality control.

compcoffee2• 8 years ago

But at least they don't seem to skimp on quality control.

Then why are so many deliveries being canceled/postponed for quality issues? Their quality control, from what I read, is horrible.

Analemma_• 8 years ago

But at least they don't seem to skimp on quality control.

I don't think this is true (remember, reworking only addresses the problems you can see).

Last weekend I rented somebody's Model 3 on Turo to see what it was like, and while I enjoyed it, the QC problems are real. The touchscreen crashed twice, and the cupholder in the center spar doesn't close unless you push it very slowly. This is in a $60,000 car! Things like that don't inspire confidence in long-term reliability and I definitely won't be spending my own money on a Model 3 until I see some assurance that they are overhauling their QC in a big way.

semi-extrinsic• 8 years ago

The touchscreen crashed twice

In this case, do you lose the speedometer etc while it resets? Does it reset automatically or do you have to press something? Genuinely curious.

Edit: this will become the go-to excuse for speeding in the future - "I'm sorry officer, I don't know how fast I was going, the speedometer had crashed and needed to reboot".

Shivetya• 8 years ago

another example was a P3D version with mismatched interior door panels. how that ever got to a customer is mind boggling.

still I have a deposit on a M3 since 8/5, will take a large number of pictures when it ever arrives and hopefully by then it comes out fine

stcredzero• 8 years ago

The battery pack with the attached electronics bay seem quite well optimized for easy assembly.

throwawaymath• 8 years ago

Is Insider an especially important source? Why is the title editorialized to include it?

dang• 8 years ago

It wasn't; it was part of the HTML doc title. Most publications do this. We take those bits out.

Theodores• 8 years ago

Seems that a lot of people think that car making begins and ends with the production line. The bit you see on TV - 'final assembly' - is not the whole car making process, it is just the coming together of the many parts that have been assembled elsewhere into bigger parts that then get put together in this 'final assembly'.

Sounds as if Tesla's 'final assembly' is not that final with a little bit of work being picked up in quality control. The spin of the story is that Tesla are cheating their figures by just repairing a few extra cars to get them out the door, but there is probably a happier story than this - at least those defects were not shipped to customers.

It is okay for a new company to take time to re-learn what has gone on before. I jokingly imagine Tesla to be actually catching up to 1970's British Leyland quality control 'circles' with the next step being a relearning of the 'Toyota Way'. It takes time to build a culture of quality control, efficiency and all these good things. The only thing is that soon our friends in China are going to be offering plenty of very nice electric cars to customers who just want electric rather than dinosaur power. Will Tesla be able to compete once China have enough EVs for their own market?

Barrin92• 8 years ago

It is okay for a new company to take time to re-learn what has gone on before. I jokingly imagine Tesla to be actually catching up to 1970's British Leyland quality control 'circles' with the next step being a relearning of the 'Toyota Way'.

This would be okay if we actually were living in the 70s right now. The problem is of course that Tesla consumes money (like basically nobody else) and the real Toyota is still around and produces as many cars on a weekend as Tesla does in an entire year.

It's not like people are going to stop buying Toyota cars out of goodwill for Tesla.

who just want electric rather than dinosaur power.

this too seems to be part of the reality distortion field around Tesla. All big manufacturers already produce hybrid and electric cars, and their research budgets dwarf Tesla, they will be ready in a few years as the market grows.

trixie_• 8 years ago

With this company so shorted how are we supposed to tell the difference between actual news and a hit piece?

ralusek• 8 years ago

Read the article and then use your noggin.

xkjkls10• 8 years ago

I think you can normally say the opposite more often. With billions invested in the company, how can you tell the difference between actual news and a puff piece. There is way more money long Tesla than short it.

zorkw4rg• 8 years ago

Its one hell of a gigantic conservative industry Tesla is up against. When jeremy clarkson (top gear) did his hit piece [1] there were (probably still are) a lot of people who were completely convinced by it and who saw nothing wrong with it at all.

I think its quite prudent to admit to uncertainty instead of being foolishly opinionated in either direction and wrong. But people are so full of themselves, its rather disconcerting.

[1] https://en.wikipedia.org/wiki/Top_Gear_(series_12)#Tesla_Roa...

melling• 8 years ago

Tesla is almost at 6000 Model 3’s a week.

https://www.bloomberg.com/graphics/2018-tesla-tracker/

And I believe that they’re building almost 2,000 Model X,S cars a week.

At almost 400,000 cars a year, Tesla has got to be doing better than many expected.

Elon seems to be doing fine in this recent interview:

https://youtu.be/mr9kK0_7x08

SilasX• 8 years ago

"Impulsively calling rescuers 'pedos'" and "impulsively violating securities laws with false claims about secured funding for going private in order to screw short-sellers" are just about the opposite of "doing fine".

Edit: the heck? How is that not evidence against "Elon doing fine"? Is this just "don't feed the trolls", or auto-downvotes of anything anti-Musk?

nickik4• 8 years ago

One slipup on twitter during an argument and and one incident where there is a question weather it is totally legal or not, lets wait for the ruling on that. Overall his companies are doing fantastic.

Animats2• 8 years ago

That's not bad for one line. Most auto production final assembly lines make about one car a minute. Tesla Fremont is now running 7 days a week, 2 shifts, and talking about 3 shifts. That's usually considered overdoing it. Few auto plants run more than two shifts, and do maintenance on the third shift and Saturday. More than that, and quality suffers and the workers burn out. 5,000 cars a week per line is about right.

xkjkls10• 8 years ago

Also, running that many shifts is a lot of overtime paid.

bryanlarsen• 8 years ago

That headline could also be read as "Tesla hit Model 3 target DESPITE reworking thousands of cars". Presumably they've got technicians both doing the rework, and also technicians and engineers working on solving the underlying issues causing the rework. Once the latter is done, they'll have freed up a lot of technicians for other tasks, and making 5000 cars a week will be almost trivial in comparison.

mcguire• 8 years ago

Which is good since other manufacturers are making 200,000 or so.

danhak• 8 years ago

It's almost as if scaling an operation > 100% YOY poses different challenges than continuing to run a stable and mature enterprise.

newnewpdro• 8 years ago

Couldn't they also be bursting to this rate by accumulating a queue of needs-rework vehicles in previous weeks, then accounting them all in the same week when the rework happens in combination with that week's production line numbers? It would artificially inflate the number for a given week, without actually being sustainable.

RayVR• 8 years ago

Not sure why this would be flagged. Seems like an article of general interest.

MBCook4• 8 years ago

I’d imagine the large contingent of hyper-pro-Tesla people who argue virulently against an any thing that makes Tesla/Elon look bad.

He’s turned into an almost political topic in that way, unfortunately.

reustle• 8 years ago

I sit in neither camp but I see almost as many anti-Tesla commenters, too

Animats2• 8 years ago

Why was this flagged?

This has been a big issue in auto manufacturing for decades. At one time, the Detroit automakers ran "dealer preparation centers" to finish what the production line hadn't done. Toyota's big innovation was figuring out how to get it right the first time, which got costs down and quality up. In a Toyota plant, if you have to stop the line for 10 seconds to get it right, a worker can do that. Videos of Toyota plants show a cycle time display and a vehicles per hour display over the line. It's expected that a car will take an extra few seconds every once in a while.

reaperducer• 8 years ago

At one time, the Detroit automakers ran "dealer preparation centers" to finish what the production line hadn't done.

Is that the "dealer prep" line that always seems to appear as a non-optional line item in the price of a new car?

rootusrootus2• 8 years ago

These days dealer prep mostly involves taking off the plastic paint protection film, removing any other hardware that protects the car during shipping (suspension blocks, door bump pads, etc), checking functionality of major components, verifying fluid levels, etc, and finally detailing the car (well, washing and vacuuming at least).

You can actually find the dealer prep checklist for a lot of cars with Google, for example GM has theirs freely published on their public web site.

jonathankoren7• 8 years ago

No. The "dealer prep" you're referring to is literally taking just taking the plastic wrap off the car.

He's talking about fixing production errors, like panels being stuck on wrong[0], or window gaskets being misinstalled[3].

As a side note, this is the first time I've heard anyone defend the lemons coming out of Detroit in the 70s. It was so bad, that my dad would check the VINs, so that he would never buy a car that was built on a Monday or Friday, because it was too likely that the workers were either drunk or hungover.[1][2][4] And this is when he was in the UAW as mechanic and then a body man.

I thought these production issues might be overblown, so I checked the deck lids on three 3s in the parking lot at work. All three were misaligned. I didn't expect them to be perfect (it's not laser perfect on my 2010 Jetta either), but I didn't expect to fit my finger through the gaps. All three cars. All three in different places. That killed it for me. I will never buy a Tesla 3.

[0] https://teslamotorsclub.com/tmc/threads/body-panel-gap-and-a...

[1] https://jalopnik.com/the-ten-worst-car-factories-in-history-...

[2] https://jalopnik.com/5645880/auto-workers-drinking-getting-h...

[3] https://teslamotorsclub.com/tmc/threads/unfinished-metal-and...

[4] https://www.thisamericanlife.org/403/transcript

colejohnson66• 8 years ago

Why was this flagged?

It seems some people flag anything Tesla / Elon Musk related

hevi_jos• 8 years ago

This is great. Mr Musk is a perfectionist, like Japanese are. He is not selling bad finished cars(like other companies will do).

They are also using lots of automation. This means there are lots of problems at first. With automation, once you solve a problem, it is solved forever.

How you could compare Tesla rework with other companies rework, where the threshold that determines if a car needs reworking is decided by the executives of the same companies?

I could make rework zero just selling broken cars. I bought a European car with a problem in one valve of the engine when it was brand new. I told the dealers about the strange sound, they told me it was normal. It was not and gave me lots of headaches later.

gamblor9568• 8 years ago

Most other companies would fire a factory manager if even a fraction of the cars produced at the factory required rework of any kind, let alone the majority of the cars requiring 37 minutes of rework.

Take your Eurocar example, and then multiply that by multiple systems. And then apply that to almost every purchaser of the vehicle. And now you understand Tesla's problem. Somewhere along the way, Mr. Musk's "perfectionism" is interfering with the actual production and quality control of the vehicles and resulting in vehicles that are very far from perfect....when they easily could be, if he were to let the professionals run the company.

sytelus• 8 years ago

The article hasn't described the root cause but Musk had hinted at it in one of the interviews. I think what happened was that when gigafactory started, they had this huge vision of complete end-to-end automated assembly. So there were robots designed for each portions of assembly, including some robots which were supposed to operate in inperfect situations through vision data, for example, locating a dangling hose, grasping it and connecting it within a car. The big learning was that robotic functions that operated in uncertain environment, relied on vision and tried to perform manipulation tasks failed big time to do the job consistently. Musk mentioned that human would pick up a gangling hose and connect to its expected position effortlessly in a second while a robot will fumble around and dig in to car even damaging it. In past few months, I think they have put a lot many more humans at work turning in to more traditional way of assembly, although more automated than other car manufacturers.

chmike3• 8 years ago

I can't stop myself thinking that this looks like short seller's propaganda. Producing a new car require adjustments in the process. This info tells me the feedback loop is in full swing to enhance production quality. I fully understand why Elon Musk wants to make Tesla a private company again.

Short seller's activity is toxic for the US industry. Allowing to make profit on other peoples failure is a good recipe to pull down an economy. Good luck with that.

threeseed3• 8 years ago

And you sound like a conspiracy theorist.

Negative news comes about companies every single day and I never hear people go on about it being caused by short sellers.

chmike3• 8 years ago

My point is that it is not negative news. It is presented as a negative and dramatic news.

slivym• 8 years ago

I think you're being a bit too sensitive. I agree with you - building a whole new production line, and Tesla's share holders understand that too. This is just more information for those shareholders. It's important that there's accurate information about Tesla's manufacturing process for the market to correctly value the stock. It's no more objectionable for short sellers to emphasise bad news to drive the stock down than it is for Tesla to emphasise good news to drive it up.

chmike3• 8 years ago

Ideally, information should be objective. The efforts and pressures should go toward this goal.

What you say, if I understood correctly, is that it is normal to have disinformation news and that they will cancel themselves out. Unfortunately, information doesn't add up like numbers.

Internal documents reveal the grueling way Tesla hit its 5,000 Model 3 target

Related Stories