GM versus Tesla: Pre-Production Comparison for Meeting Model 3 Deliveries At 2017’s End



All car manufacturers test their vehicles prior to production. Tesla prides itself in taking a Silicon Valley approach, shortening the procedure where ever they can by making the process more efficient (ref 10).

Tesla Model 3 for 2017? A look at some comparables

Tesla Model 3 for 2017? A look at some comparables

The biggest game changer for Tesla was over the air updates-OTA, in which Tesla can update vehicle software without having customers bring vehicles into their service center, saving them time and money. OTA’s also allow Tesla to do some late stage testing in the field via their customers.

GM has also finally started using OTA updates (ref 8) on their Chevy Bolt EV, so Tesla no longer has a leg up on OTA updates in this case. In addition to OTA updates, Tesla does its testing in house to speed up the iteration process, but it is not clear if they do more in house testing than GM, as GM also has extensive in house test facilities also.

One thing that both Tesla and GM have in common is that their pre-production testing is comprised of three parts: mule testing, early prototype vehicle testing and, in part 3 they test their production line….and of course the vehicles that get produced on the production line.

Let’s look at the details and how long all this takes.

One of the first production Chevrolet Bolt EV's roll off GM's Orion, Michigan assembly line

One of the first production Chevrolet Bolt EV’s roll off GM’s Orion, Michigan assembly line

GM Pre-Production Testing

GM has some complicated names for their pre-production vehicle and manufacturing testing. For those interested in the details reference 2 and 3 (below) are a good read.

General Motors’ pre-production Bolt EV test vehicles – IVER, or integrated engineering vehicle release, numbered between 55 to nearly 100 (depending on source – ref 2 and 4).  These vehicles were tested for approximately 1 year following mule testing, and were first spotted 1.5 years prior to production start.

According to Forbes (ref 4):

Early Chevrolet Bolt "mule" in action (January 2016)

Late stages Chevrolet Bolt “mule” in action (January 2016)

“These integration test vehicles include virtually all of the parts that will go into later production examples and are generally built from production tools. These vehicles usually appear 18 to 24 months before production start. In the case of the 2017 Chevrolet Bolt EV, almost 100 integration test vehicles were built starting in early 2015.

These integration test vehicles go through a staggering array of tests to make sure everything works as intended and continues to work over the expected lifetime of the car. For most modern cars, specifications call for most parts other than wear components like tires and wiper blades to survive 10 years or 150,000 miles. To ensure that happens, cars are run through foot-deep baths of salt-water hundreds of times, passed through car washes 500 times, exposed to hurricane-force rain chambers with water pouring down at 2,000 gallons per minute and driven tens of thousands of miles over cobblestone and dirt roads to try to shake everything loose.”

One of the first PPV Bolt EVs validating the process in March 2016

One of the first PPV Bolt EVs validating the process in March 2016

GM’s manufacturing validation of vehicles made on the line (and also testing of all the production line machines) starts 6-9 months before formal production. It is broken into 2 phases: PPV (or production process validation) and MVB or (manufacturing validation build).

For the Chevrolet Bolt EV the entire pre-production testing and manufacturing validation lasted 2 years from the date they first showed the Bolt EV concept vehicle. Quite a feat since the Bolt EV had an all new battery pack, power train and body.

When GM fired up the line, they had an initial production rate of 9 vehicles/hour (ref 5) = 360 cars per week. Faster than Tesla’s initial production rate on both the Model S (15-20 cars/week, ref 15) and model X (39 cars/week, ref 12).

Bolt EV Pre-Production Testing took only 2 years

Bolt EV Pre-Production Testing took only 2 years

Tesla Model S and Model X Pre-Production Testing

Pre-Production Tesla Model S Being Assembled In Early 2012

Pre-Production Tesla Model S Being Assembled In Early 2012

Tesla pre-production testing is also broken into 3 stages just like GM: mule testing, prototype testing of vehicles (that look like the production vehicle but are not made on the final production line, referred to as “Alpha” vehicles), and vehicles made on the production line referred to as “Beta vehicles” (ref 9,10 and 11).

Tesla’s Model S alpha test fleet consisted of approximately 100 vehicles and, like GM, are put thru various and tortured testing to validate the design. Alpha testing lasted nearly 2 years for Model S.

Beta testing of Model S came next. The entire pre-production time line for Model S start to finish was approximately 3.25 years and initial production rate was 15-20 cars per week.

Tesla Model S Pre-Production Testing lasted 3.25 years

Tesla Model S Pre-Production Testing lasted 3.25 years

Model X pre-production testing lasted 3.6 years including all the delays due to problems with the falcon wing doors and motor over heating issues while pulling a trailer (wiki).

Model X Pre-Production Testing took 3.6 years

Model X Pre-Production Testing took 3.6 years

Tesla Model 3 Pre-Production Projected Time Line


First Model 3 prototype broke cover on March 31st, 2016

In late 2015, Tesla announced that most of its engineers were already working on the Model 3, not Model S or X ( ref 16). The prototypes were unveiled on March 31, 2016 and the number of reservations for Model 3 exceeded Tesla’s imagination.

As such, Elon Musk decided to expedite the program, telling suppliers they needed to supply production parts in around 1 year or 6 months prior to 1st prodroduction vehicle target date (ref 20). Coincidentally, we can note that this lines up well with what GM uses for production line testing so it’s not out of line with accepted practices at GM.

Tesla Model 3 Projected time line is 2.25 years

Tesla Model 3 Projected time line is 2.25 years

Also note that the Tesla Gigafactory is targeted to manufacture cells by end of 2016 (ref 22) but as of yet we have not heard of that startup, but even if that date slips somewhat it is still around 1 year from 1st vehicle production – so not a show stopper for getting first delivery about one year from now. Also note suppliers still have 6 months to start delivering production parts.

Let’s look at how the Tesla Model 3 timeline compares to the Chevy Bolt EV and the Model S and X.

Model 3 Pre-Production timeline is not out of line compared to GM Bolt EV

Model 3 Pre-Production timeline is not out of line compared to GM Bolt EV

Can Tesla still make their target date of end of 2017?

Reasons they will make it:

  • Timeline is tough but not out of line with what GM did on the Bolt EV
  • Supplier deadline is 6 months prior to 1st delivery, in line with GM practices

Reasons they won’t make it:

  • Tesla consistently misses their delivery dates
  • Suppliers could be late delivering parts
  • GM and Tesla’s Model S had visibly more cars testing than Model 3. In the case of the Bolt EV we had sightings and articles describing a test fleet of 55 vehicles a full 1.5 years from production start. In Model S case we had articles written and test vehicles sighted a full 1.75 years prior to production start. Here we are 1 year away from production start for model 3 and  how many reported sightings of test fleet have we seen? Close to zero.

What do you readers think? Will Tesla make their bogey or not?

About the author: George Bower is a retired mechanical engineer with over 20 years experience in gas turbine power systems including the design and development testing of the auxiliary power units for both the A320 and Boeing 777 commercial aircraft. He leased a 2012 Chevy Volt for 3 years, and currently drives a Tesla Model S.


1)IEV’s: 55 Bolt EV pre-production test Vehicles
2)GM Inside News: GM prototyping stages
3)Electrek: Chevy Bolt EV prod intent vehicle spotted
4)Forbes: How GM is testing the Bolt EV
5)Wikipedia: Chevy Bolt EV
6)IEV’s Chevy Bolt production underway
7)Detroit News: GM readies for Bolt EV production
8)Electrek: GM OTA updates
9)Model S Alpha fleet
10)Tesla Testing explained
11)Autoblog: Model S 100 test cars
12)Wikipedia: Tesla Model X
13)Wikipedia: Tesla Model S
14)Wikipedia: Tesla Factory
15)IEV’s: Tesla Model S sighted
16)Wikipedia: Tesla Model 3
17)Hybrid Cars: pencils Down
18)Bloomberg: Pencils Down
19)Geekwire: Tesla shortens up Model 3 schedule
20)Electrek: Tests sets deadline for Model 3 suppliers
21)Electrek: Tesla finishes production line layout drawings
22)Cleantechnica: Tesla giga factory to begin

Categories: Chevrolet, Tesla

Tags: , , , , , , ,

Leave a Reply

121 Comments on "GM versus Tesla: Pre-Production Comparison for Meeting Model 3 Deliveries At 2017’s End"

newest oldest most voted

Good sumery, and i love the reference links 🙂
The lack of model3 test fleet is worry some, i hope they rather be abitt late than shortening testing and risk recalls :S

Could they have disguised them as Chevy Bolts? 🙂

Tesla would never make anything so ugly even intentionally.

Have you not seen the Model X? I much prefer the looks of the Bolt EV.

Comparing Tesla & GM is Like comparing a 100yr. 0ld Man to a an 8yr 0ld Child…While the 0ld man is tired can’t be bothered & SET in his ways,the 0ld man Avoids, Denies , Resists, and Refuses Innovative Ideas & New technology with all the strenght that is left in him…..Meanwhile , the Child is full of Life Eager , Curious , Anxious to discover & find innovative new ,Safer, Better & Cleaner ways to Transport People & things , As the 0ld Man(GM) Withers Away…

Except, um, 8 year old children aren’t generally good for squat in terms of producing anything…basically no different than 100 year old dudes.

p.s. the Bolt is in showrooms on time and as promised…so, um, Old Dudes FTW!

A lot of Nuts & Bolts in those showrooms…lmao

Yeah that old man just cranked out a car in 2 years flat. That old man has a lot of resources while the 8 year old might be breaking his piggy bank to see if he can count his coins enough to get a model car.

The promise was in 3,000 showrooms in 50 states.
I know I have to take my socks of to count further after 10, but a handful of showrooms in two states? The Bolt’s launch is starting to look worth than the Model X launch.

Not really, dealers are already receiving large volumes of cars. The limited launch is to hottest markets (and of course to maximize credits for the quarter).

Actually, GM’s official stance was to deliver the cars to select dealers by end of year (once last year one GM person said they were planning on 50 state rollout, but all this year they have stated select dealers).

There are already hundreds of bolts available to dealers and I imagine by the end of the month they will have delivered in two weeks as many as Tesla did Model X units in 5 or 6 months.

I gave $1000 to an 8 year old lol

LG does the battery testing.
But, GM’s 40% loss of capacity? Did they push the LG chemistry too far? And, does LG not want the warranty work?

GM warranty is for 40% degradation and they will replace the battery if it falls below that.
Tesla specifically excludes degradation under their warranty so a battery with 50% (or even more) would not be covered under warranty.
Only Tesla, Ford, Fiat and Mitsubishi specifically exclude degradation – every other BEV maker provides a degradation warranty

Yeah, don’t read too much into it. Great analysis on this article!

Great info to know…
Judging from how excellent LG batteries have performed so far it appears to be pure CYA…

I think they have to do it in the timeframe they’ve already announced. In the past they were something different, something new, and no one expected much from them. That currently isn’t the case with so many preorders and very high expectations. I very much think they will hit their production dates. They are at the sink or swim point with the Model 3, and I not only want to see them swim, I want to see them fly! Go Tesla!

Agreed, those analogies with Model S and X always forget That Tesla is still in infancy, learning to walk, now they can start to run… and fly!

I believe Tesla will make it, however production numbers will be tiny.

Tesla will make it just fine…It a product worth waiting for..

I thinks so too,It could be similar to model
L x where 2months worth of production or 300-500 cars by end of 2017. But how knows I wouldn’t be surprised if 5000+ are delivered in Cali.

Even established OEMs usually struggle to hit their SOP (start of production) dates. Even when they manage to meet them there is a ramp-up curve and production output is usually very small at the beginning. You usually see an inherent learning curve for building a completely new vehicle with both suppliers and the OEM so the production line moves at very slow pace, initially Just look at the Chevy Bolt, our the i3 when it came out. Even though the vehicle is officially launched, barely any vehicles reach the dealerships, bar the final customers in the first 2 months.

Lack of visible test fleet is a big indicator that they have not built the Beta vehicles yet. From what I can tell, the only cars spotted have been the early prototypes that were shown at the announcement party. Please correct me if that is not the case (with references, of course).

If Tesla chooses to “compress” timing by shortening the limiting validation testing, they are laying the ground-work for all kinds of quality issues once they hit volume production. Trust me, you don’t want to find a previously unknown failure mode when you’re cranking out cars at a rate of 2000+/week.

tesla is a good marketing operation; tesla will get *something* out. in the worst case, i can see them releasing a limited number of cars to “ambassadors” – much as bmw did with the “active e” program. these cars would essentially be prototypes that would be driven by select tesla ambassadors. this would give visibility to the model 3 and provide “proof” that delivery is near. tesla will manage this event with its usual adept usage of social media to keep their fan enthusiastic.

JeremyK said: “Lack of visible test fleet is a big indicator that they have not built the Beta vehicles yet. From what I can tell, the only cars spotted have been the early prototypes that were shown at the announcement party. Please correct me if that is not the case…” I’ve Googled [tesla model 3 spy photos] and it does indeed seem that all there are, is the (as I recall) three prototypes shown at the Model ≡ reveal, and used for test drives. Or maybe only two of those; I’ve seen at least one claim that one of those was not drivable. This does indeed seem to point to a delay in Model ≡ development. I won’t claim it’s definitive; as Carl Sagan famously noted, “Absence of evidence is not evidence of absence.” But it does at least appear to indicate that Tesla isn’t progressing as fast in its development of the M≡ as its aggressive development schedule suggests it should be. But then, personally I’ve been predicting that Tesla will be about 6 months late in getting the car into production. That would still be much less of the approx. two year delay they had on the Model… Read more »

Not only a lack of test mules, but we haven’t even seen the production vehicle. How long ago did they say “pencils down”? It’s almost 2017 and they haven’t even shown the final design.

I think Tesla will make it.

Not only does the Model-3 seems optimised for (relative) easy production, other then the Model-X the stakes are also very high to reach that timeline.

Note that Tesla might be able to do many more testdriving in-house compared to ICE as EV’s do not polute the air and Tesla has very good reasons to keep certain features under cover (in order not to ‘Osborn’Model-S sales).

This. Tesla is quite capable of doing the vast majority of testing indoors. We already know they have an indoor test track, and the weather abuse chambers are already indoors, and so on. They have incentive to do all that testing indoors to avoid Osborning.

Well…”In late 2015 Tesla announced that most engineers were already working on Model 3, not Model S or X ( ref 16). The prototypes were unveiled on March 31, 2015 and the number of reservation for Model 3 exceeded Tesla’s imagination.”…Actually, the TESLA Model 3 was revealed to the public this year: March 31st, 2016!

That the author had so many detals but missed that Date Typo suggests HE did not Line Up on March 31, 2016 to put His Reservation in! (Or, I am pretty sure that date would be better planted in his mind, more acurately!)

Also, how could they reveal the Tesla Model 3 on March 31, 2015, BEFORE late 2015, when only then did Tesla state most Engineers were working on in?

Hey Robert,

I think was just a simple mistake/typo in the text in that one spot (now corrected).

As you can see from the chart, the debut of the Model 3 was correctly placed in 2016, and all the surrounding data work is in reference to March 2016 being the debut date. Sometimes hiccups happen, I don’t think we can draw deeper conclusions other than that.

Besides Testing Model 3 inside the plant, some could be hand built, boxed up (put into fully closed shipping trailers), and shipped out for testing in Special Testing Facilities, like the cold weather climate testing facility in Florida!

Some could be shipped for hot weather testing to Australia, there are lots of roads and places there with few people and less traffic!

There are a lot of places where some of those supposedly 300 cars worth of parts could be built, and being tested, quietly, without a lot of prying eyes. Even Africa and Mexico and The Yukon Territories in Canada, all come to mind!

They are not restricted to doing all their testing within 50 miles of the Fremont Plant, in front of curiosity seekers, spies, and other competitors!

somebody here seems to be severely geographically challenged when he can’t tell the difference between a continent, a country and a province…

just to update you, the *continent* (not country) of africa has changed a lot since the films of martin and ossa johnson in the 1930’s. believe it or not, there are actually *cities* in african nations. there are highways too. yes! it really is true!

Your complaint is misplaced. Robert said nowhere in his comment that he was specifically referring to continents, or countries, or provinces. You may have inferred it, but he didn’t imply it.

What Robert suggested seems rather unlikely to me — don’t car makers actually want “spy photos” to be taken of their new cars, to create media buzz? — but there is nothing wrong with his geography.

i wasn’t complaining, i just think it was a stupid comment…but you might want to check your geography also if you missed my point.

Are you really sure there is a point for you to make?

yes, i am. it was a ridiculous comment.

The author missed an important part of the Tesla timline.
Building 100,00 – 200,000 cars on a 400,000 a year productionline in 2017.
The suppliers are not ordered to deliver production quality parts in 2017-07 for assembly of beta-test cars.
They are expected to start delivery for mass production on.

The order for parts for 300 cars are the parts for alpha and beta cars.
The manufactoring validation of vehicles made on the line (GM phases PPV and MVB) is planned for the first half year of 2017.
I expect it mainly in Q2-2017.

As Elon said, the ramp up is hard to predict, but at the start of 2018 the production will be at 400,000 per year or 8,000 per week.

Tesla’s official goal is run rate of 400k Model 3s per year by the END of 2018.

The stated intention of Tesla is to produce 100k – 200k IN 2017 and to produce 400k IN 2018.

2017-07-01 is the intended start of mass production with a ramp up to 8k per week. Because the speed of the ramp up is an unknown, the forecast for 2017 is width.

The writer of this article is a half year of on the end of the Model 3 timeline.

I do not know if this is due to a honest mistake, poor reading skills or a prejudice against the new kid on the block. But he should plot again his timeline with the correct end date.

You really think Tesla is going to hit the 200k/annual rate in 2017? When they haven’t even begun testing Pre-production vehicles? ? I think the author is right, it appears as if they will deliver a few in 2017 and ramp up to a high level by the END of 2018. B/c at the beginning of 2018 they will still be working out the bugs, working to increase production and possibly hitting their stride, by mid-year.

Twelve thousand cars in 2017!

“The stated policy of Tesla….”

You are mistaken. Or lying. Tesla has repeatedly stated their goals are:

1. Production in late 2017
2. 400k unit/year run rate at the end of 2018

Musk has made random throwaway comments that contradict this official schedule. Nobody, inside or outside of Tesla, pays any attention to his random comments. Except, apparently, you.

Have you ever heard of a nasty little committee called the SEC in New York?
CEOs are not allowed to make little random throw away statements on things like this, and they make sure they don’t.

SEC, lol. How many times did they prosecute MSFT for missing every single Windows release schedule by 1-3 years?

This Q&A was covered by the standard “forward looking statements” clause. Furthermore, it was clearly a mistake (he gave the 2H18 number but the question was 2H17).

I believe the Model-X delay was largely due to the falcon doors. So, if they keep things more traditional with the Model-3, hopefully they can get it done faster.

“I believe the Model-X delay was largely due to the falcon doors”

David, why were there even more problems with body panel fit, interior trim (like carpeting) terribly flawed, missing fasteners, huge internal joining gaps, failures by seat suppliers to execute, fabrication problems with glass, malfunctioning front door (NOT FWDs) presentation and closure, etc etc.? And the MX had significant commonality with the MS, as stated by Straubel.

It’s more than doors. The MX was a messy and very poorly executed development, period. Go back and spend 30 minutes at the Tesla Investor Relations page and skim quarterly shareholder letters from mid-2014 onwards. Slip after slip despite frequent assurances that the schedule was “accelerating”.
BTW: on the subject of doors… does anybody else wonder about the settlement with Hoerbiger? Wonder why that whole event shrunk into the darkness? We’ll have to wait for the 10K in February to ferret out the cost of the settlement but it appears that Tesla had to sign a check to the mean old Swiss company that dared to defy The Elon.

Model 3 won’t hit the streets with actual production cars until mid-2018.

Why would Tesla write a check, they are the ones that brought the suit?

“Tesla and Switzerland-based supplier Hoerbiger Automotive Comfort System have agreed to settle a lawsuit—filed by the electric carmaker—over problems with a proposed hydraulic system for the Model X’s falcon wing doors, reports Reuters. The details of the settlement, however, are not yet public. Both parties have until 13 October to file a joint statement.”–Reuters

Ff: ff: Tesla sued because Hoerbiger asserted they were owed money for delays and other issues in keeping with the termination clauses in the contract. Tesla disagreed – vehemently enough to sue Hoerbiger for lack of performance and refuse payment. Tesla changed representation during the case and then a settlement was announced. My belief (and that’s all it is now, but I’ve been in a couple of these on both sides) is that Hoerbiger had case and Tesla had to back off. Your belief is as valid as mine right now, but indications are that my guess is right. Once the 10K comes out I’ll post here either to swagger or look at my shoes.

I said Tesla sued Hoebinger, which is not a belief it is a fact.

Relax: I never said they didn’t. What I’m telling you is that Tesla only filed a lawsuit after Hoerbiger demanded to be paid what Tesla complained was “a large sum of money” per the contract, which Tesla believed had been breached. Now it appears that Tesla is in fact obligated to pay the amount that they sued to avoid paying. That is my belief. Tesla sues a lot of people and generally doesn’t do too well (ask Jeremy Clarkson).

Sure, I think that fits my remembrance. I don’t need to relax as I am not upset. Tesla also gets sued by a lot of people, not sure on how litigious they are personally, and I really can’t take your word for it, in comparison to what, by what measure. Usually in cases such as this lawsuits and counter-suits are flying back and forth.

Tesla ended up paying substantially less than Hoerbinger had asked for, if I recall correctly.

Not to weigh in here with an opinion as to who is right or wrong here, but a bit of additional info:

Jay Cole reported that Tesla only asked vendors for a proposed design for pigeon falcon wing door actuators a very few months before production was supposed to start, so Tesla didn’t really give vendors sufficient time to develop and test them; less time than the industry standard. That would certainly have been a major factor in any contract dispute, arbitration, or lawsuit.

Hire Useless Liars to do a Job for you That’s what happens …The Falcon Wing doors are Working Properly Once they Hired Competent people of Value!

David Murray said:

“I believe the Model-X delay was largely due to the falcon doors.”

At one point, Elon claimed they had even more trouble with the “captain’s chair” second row seats. Dunno if that’s true, but it’s what he claimed.

Everybody talks about problems with the FW doors, but that doesn’t equate to proof that it was tthe biggest contributor to delays in Tesla getting the MX into production

Good article George. I think Tesla will do exactly what GM has done with the Bolt regarding the Model 3 launch. That is to say, very late in 2017, Tesla will produce a very, very small amount of Model 3’s, just to say “we did it!” And after that, they will flow out of the factory at a trickle for probably a year until they work out the bugs with the car and the factory.

Indeed we all want attractive, capable, and affordable EVs as soon as possible, but we also want then to be high quality products. Considering the Model X release, I hope that Tesla take their time with the Model 3 to get it right.

As always this is a great analysis by George Bower

There’s one issue I see with the Model X timeline. It notes a test vehicle sighting in January 2014, but all indications are this was the same concept vehicle that made public appearances starting at NAIAS 2013 up through CES 2015. It wasn’t until early 2015 that true test vehicles started to appear, approximately 8-9 months before the final reveal and deliveries.

If Model 3 is going to be on schedule, there needs to be an increase of new test vehicle sightings (likely with some form of camouflage) in Q1 2017. So far we haven’t seen any other than the ones shown last March at the concept reveal.

Why? The testing can be done indoors.

Tesla’s trying to conceal the real interior (Which clearly wasn’t revealed at the first reveal), and is trying to avoid Osborning the Model S and X, and is known to have indoor test facilities.

Is it confirmed that the Bolt can receive OTAs?

I thought one of the reasons reviewers lamented the fact that the regen setting must be selected every time you enter the car is that the car couldn’t be updated to fix this oversight…

Yes….though probably no system-critical updates. Think more like updates to Android auto/Apple Car play… recalls will be pushed out via OTA.

I don’t expect that to change. It’s the same as the Volt or Leaf in that way.

The manual says the infotainment system can get OTA updates. That might mean the entire interior gauge system if we’re lucky. It’s a big less likely that anything drivetrain-oriented will get them.

What no one seems to have mentioned here is that the first few thousand people in line for Model 3 reservations are Tesla employees. They will be beta testing those Teslas for us.

Major parts suppliers don’t need their parts “beta” tested. Their parts are unit tested at the factory.

The only thing you need near production is “integration” testing, and you can do much of that via computer modeling.

Reality check: Legacy auto makers, including GM, do extended test driving — for up to six months — of early production units, before selling any new car model to the public.

Tesla doesn’t, probably because it’s a new and growing company that needs every dollar it can get to fund its growth; a company which literally cannot afford to wait those few or several months before it starts selling a new model of car.

Surely Tesla is aware of the fact that it could avoid many of the problems reported online if it would spend some time doing extended testing after start of production. Most likely Tesla continues this strategy out of necessity, not choice.

We hold this truth to be self evident that, unlike men, not all ota updated are created equal. I almost stopped reading when I read that casual comment that Tesla’s lead in that area no longer exists. Bah Humbug!

In regard to whether Tesla will make their production targets it’s just speculation. Their original target was 2020 and people then were even saying they would not make that, then they upped it by 2 years, so that will be quite a feat if they succeed.

Not that 2020 bullpucky again. That’s already been debunked. It simply doesn’t exist. That false rumor comes from a target date for manufacturing half a million units combined of models S, X, and 3 together.

There was never a target date of 2019 or 2020 for volume production of the M3. And we’ve got sources dating all the way back to 2013 that all say 2017, with new source each year, every year since then that all say 2017. Prior to 2013, the target for what they then called GEN III was actually earlier than 2017.

How about we kill this fake news meme?

It matters little now since that went by boards a long time ago.

It does matter, since Tesla bashers are still trotting out the claim that the original target date for production was 2020. As Nix said, that’s completely untrue.

However, I think Nix exaggerates when he says that Tesla’s stated goal was always 2017. There are reports from earlier years of Tesla spokesmen giving other years as the target for start of production or deliveries. I see that Elon once Tweeted “2016 or 2017”, and the 2016 date was given elsewhere:

Yes, there were earlier dates than 2017 for “Model E” and “GEN III” and “BlueStar”. Such as that link you provided. Never later dates though, even back then.

But ever since they announced the official name was “Model 3”, every reference to a date has been 2017.

I don’t know, pointless to argue. I think it was just that by 2020 we will have 500k Tesla cars on the road, then people said well then some of those will have to be Model3, but then I recall when news first hit about when the 3 would actually dropped, everyone was surprised saying they moved up the production date.

I don’t understand why the “start date” for the “2 years” Bolt development is when they showed the first concept car, yet the “start date” for the “2.25 years” Model III development is an article saying “most Tesla engineers are working on Model III.”

Shouldn’t the apples-to-apples start be at first prototype reveal?

There is no true apples-to-apples, it is all just very rough estimation anyways. Because there is nothing that says both Tesla and GM showed their concept cars to the public at the exact same point in their development cycles.

“George Bower is a retired mechanical engineer with over 20 years experience in gas turbine power systems…”

So I’m guessing Allied-Signal/Honeywell then? Congrats on two successful programs. I was a frequent contractor participant with Sundstrand/UTC over roughly the same period, so I’m guessing we probably locked horns from time to time.

I’m certain you have many memories of transition from breadboard to “brassboard” to Iron Bird to safe-for-flight for testing (but not certified) to Fully Certified designs. If you’ve been involved in the automotive side from the Garrett/Honeywell processes, you know that the automotive requirements for production readiness are even more rigorous in many ways.

Do you seriously, honestly believe that Tesla is even close to being on a track of rigorous manufacturing process prove-out that your buddies from the Turbocharger business have to meet?

2017 isn’t going to happen.

BTW, your chart of the M3 development has one other serious error: Tesla leadership asserted 100,000-200,000 Model 3 produced in 2017 and 500,000 deliveries in 2018. Your chart shows that the plan was for “first delivery targeted 2017”. Not the case – if it is the plan today Tesla should update that.

Elon has made it very clear of their plans to change the auto industry. Investors who are squeamish or looking for a quick return have missed the window of opportunity for the moment. Technically speaking the original timeline was longer before the rush of interest in the Model III.

This article is a well thought out presentation of data from a very bright retired engineer. It is not quickly thrown together article. No need to over analyze the fine points for as George concluded, we won’t know for sure until it gets here.

My armchair guess is the same as Schmeltz and others that they will work hard to ship a few units just like the Bolt by the close of 2017, start a slow production aimed at the CARB states + fully loaded orders and maybe make full production late 2018.

If investors have a problem with the CAPEX don’t invest. There is risk here. I am not crazy enough to invest more than I can afford to lose, but I can’t think of a single company postured to change the world more than Tesla. For that reason, I’m in.

“This article is a well thought out presentation of data from a very bright retired engineer.”

Actually, it’s a rather odd combination of solid facts explained well, and arbitrarily chosen guesswork (most notably, the dates given for pre-production starts) which unfortunately are presented as facts. There are also a few outright factual errors, as noted in other comments.

This could have been an excellent and authoritative article, if more time had been spent fact-checking and if the difference between facts and speculation was made clear.

Wow, You are saying it way better than I did in poor attempts.

Hovis, at no point did I pick on Mr. for Bowers for his post. As you may note in my response to him, I congratulated him on his experience with A320 and B777 APU projects, which are no mean feat. But there are some serious errors in his narrative that genuinely miss the mark. For example, this: “Supplier deadline is 6 months prior to 1st delivery, in line with GM practices” That’s completely nonsense, to be blunt. “Supplier deadlines” in GMs current system are to have built products not just off tool but off process (that is, with production tooling and processes), completed all required environmental and endurance testing at the component level (such as water impingement, for example), with all PPAP documentation submitted. If required, all hardware/software integration testing and demonstrate full compliance with safety-critical validation requirements of standards like ISO26262. There is simply no way that a “pencils down” to the Tesla requirements definition in May of year zero supports production in July of Year 1. This is nothing like GM practices (or Daimler or Toyota or anybody else in rate production of automobiles). I don’t question his aero experience or his good intentions in providing this post.… Read more »

That sounds like a lot of silly paperwork, given that GM still ended up with faulty ignition switches.

Never forget that Tesla is bringing massive amounts of manufacturing in house. Their suppliers are supplying at a lower and lower level — they’re often buying screws and bolts, not entire assemblies. It’s a lot easier to get validated parts that way. The bigger assemblies like brakes are almost off-the-shelf.

They weren’t faulty. You’re just not supposed to hang an anvil from your keychain.

Realistic – Tesla’s official position is a 400k Model 3 run rate and the END of 2018. Look at the earnings release and Musk’s prepared statement at the start of the conference call.

Later in the call, Musk responded erroneously to a question about second half 2017 Model 3 production. That’s when he said 100-200k units. I don’t know if he misheard it as 2018 or if he was just lost in space (not unusual for him during Q&A). Whatever the reason, everyone immediately ignored the statement as an obvious error. The only people who even remember it are TSLA bears playing a silly game of “we caught Elon lying, nya nya nya nya nya nya”.

I also see they are spreading the fake news of losing $9k per car.

Tesla (Musk or Straubel; I don’t remember which) has already said they’re testing the M3 driveline in Model S chassis. I expect that the MS and MX will switch to the new M3 inverter, cell and battery pack basic design, and perhaps even share some motors — i.e. M3 rear in MS/MX front position. I would bet all these components are already under extensive test in an internal MS fleet. There’s still a lot of production validation work to be done, but there’s a reasonable chance that M3 production will be robust by early 2018.

rashomon, that’s the most rational optimistic position on M3 posted here, and I think you’re probably right about the power electronics for sure. There is far more to making a successful high rate production car than the drive train, and by all means Tesla sorely needed an update to the inverter, but this is a place that progress is likely happening.

Great article. You can have an opinion was to whether you think Tesla will hit their launch target for the Model 3 (and of course the numbers), but with this article it’s at least an informed opinion.

One factor not mentioned – GM was able to retool an existing line. Tesla has to build one from scratch plus hire and train personnel to run it. The latter certainly has more potential for delay.
GM had the line up and running 100’s of pre-production units in March. Tesla has some real challenges in the next few months if they have any hope of keeping to their timeline.

Gary, “One factor not mentioned – GM was able to retool an existing line. Tesla has to build one from scratch plus hire and train personnel to run it. The latter certainly has more potential for delay. GM had the line up and running 100’s of pre-production units in March. Tesla has some real challenges in the next few months if they have any hope of keeping to their timeline.” GM is actually at a bit of a disadvantage because of that. The Orion factory was being used to build Chevrolet Sonic’s and the Buick Verano’s before it was used to build the Bolt. That means GM had to build a parallel mock-up assembly line in order to test out their Bolt assembly process. Then once they were done with testing, they had these additional tasks: 1) Schedule stop of production of the old cars 2) Take out the tooling for those old cars 3) Take the tooling from the testing line, and reinstall it into the actual production assembly line. 4) Restart the line and validate the line is correct. Traditionally, that shutdown and restart has taken from 2 weeks to 9 weeks. Since Tesla is starting with a… Read more »

I like Tesla, but I don’t think they’ll make it that soon, and I sure wouldn’t buy an early production model 3. The lack of a preproduction test fleet is worrisome. If anything, they probably need longer to test and address quality issues simply because they’re a new and small manufacturer. Shortening the process will almost surely lead to quality issues of some kind.

John, unless you work at Tesla or currently own a Tesla, and live in California, you won’t have a choice as to whether you want to buy an early production unit anyways. You can’t.

That solves that little dilemma.

Great article George.


Tesla will be late…….business as usual.

Obviously they can show a near production car and deliver 12 to Elon and his friends by the end of 2017. But to have real production by the end of 2017 appears close to impossible.

The time lines in this article do not line up. The timelines are certainly not an apples to apples comparison. The article presumes that GM started testing the Bolt EV when the concept was launched 2 years ago. Which is certainly false. GM would have started testing powertrain mules about a year before the concept vehicle was launched. The article shows a Bolt EV “mule” in a picture from CES. That was most certainly not a mule. This was an IVER pictured and giving test drives. Bolt EV design freeze likely happened sometime in mid-2015. Mules are Frankenstein cars designed to test new components before design freeze. IVER’s come after design freeze. Also keep in mind full product design of the Bolt EV started sometime BEFORE 2013. As for Tesla we have not yet seen an IVER yet for the Model 3. The Model 3’s shown to date where more or less show mules or rolling concept cars. They certainly were not production intent as the design was not yet frozen at that point. In fact it’s rather disconcerting that we haven’t even had a spy photo of a Model 3 IVER yet. So unless Tesla is testing these… Read more »

Just of note below is a picture of a “Mule”. This was the test mule for the Gen I Volt using a heavily modified Malibu for testing.

Thanks, Nero. Now that’s a mule. It’s the mating of new stuff on an existing platform (or vice versa, a new body concept with some existing systems installed to make it work). Like a horse and a donkey, it’s made to be worked hard. But it’ll never reproduce. Thus a Mule.

Of course the awesome red M3 at the Gigafactory isn’t even a mule.

“GM would have started testing powertrain mules about a year before the concept vehicle was launched.”

We actually don’t know when Tesla started powertrain testing in mules. A Model S or Model X mule with an M3 drivetrain could look just like any other Model S or X from the outside. All we know publicly was when the Model 3 concept vehicle was launched.

Q: What does “IVER” mean?

A: IVER stands for Integration Vehicle Engineering Release (prototype development vehicle)

So how does that compare to a “mule” or a “pre-production unit”? I confess the term “prototype development vehicle” doesn’t mean anything more to me than simply saying “prototype”.

An IVER is a production intent car built as per design freeze. GM for example builds it’s IVER’s out of a special low volume facility in Warren, MI. These cars are not necessarily built on the final assembly line and some of the tooling might be prototype vs. production tooling. Note just because a vehicle reaches design freeze doesn’t mean that the design can’t change. Though there will be likely no major changes. Just tweaks.

pre-production is simply a generic term for any car manufactured before production.

A mule is something that happens before design freeze. You take components and you modify other cars to test new components that will go into a new car. Mules are often not pretty as they are often butchered to make things fit. The Gen I Volt for example had at least a couple different kinds of mules running around (the pictured hacked Malibu in my post above) and there were some modified Chevy Cruze’s that were used as mules as well.

GM’s special prototype factory is yet another difference between how GM does things and how Tesla does things.

Tesla’s fleet of around 300 test vehicles (or at least substantial sub-assemblies) will be built on the actual final production assembly line. An assembly line that Tesla has been building out since late summer, and they have been sub-assembly testing since October. (ibid on links, posted previously)

This is reasoning by analogy. Tesla has every incentive to do their testing indoors.

How do you put a million miles on a car testing it indoors? I think this simply demonstrates why Tesla’s have so many problems. You can’t validate computer models with cars indoors.

this is a well constructed article. even though i am always skeptical of wikipedia as a source, i very much appreciate that the writer included his references.

Wikipedia, like all encyclopedias, is a good quick reference but is a secondary source of info, and thus cannot be considered authoritative. For those looking for deeper knowledge, Wikipedia’s list of references is often an excellent source for primary sources on any given subject.

The problem is that a lot of people who know even less than what Wikipedia says about a subject, post as if they are an authority on the subject. So that is why you’ll so often see Wikipedia cited in Internet posts. It’s often a more polite way of saying “Dude, even Wikipedia knows that what you’re saying isn’t true.”

wikipedia is *not* the same as a conventional encyclopedia. there is a review process for evaluating information that will go into an encyclopedia. so you have some assurance of the integrity of the information published in an encyclopedia. wikipedia, by contrast, is an *open* reference, so anybody can put stuff into it without any kind of review process. so you can never trust that stuff in wikipedia is correct.

As others have said, hitting the launch date may be the easy part…it’s ramping up that will be hard and the longer it takes, the longer the competition (who already know how to build and distribute cars in volume) have to close the gap By “competition” I’m not referring to the Bolt either. I am talking about the Germans. At some point Audi, BMW, Mercedes, etc. deploy a decent electric car or a plug-in hybrid with more serious range and at that point, a lot of buyers will likely consider those alternatives.

To “compete” Tesla has already clearly moved on from “electric” as their primary motivator to “autonomous” (look for the new InsideAVs website soon). Indeed, as a long time car guy I am amazed how all of the car magazine articles have already moved past talking “electric” and are all about autonomy now. It is one area where Tesla still holds a lead, but will any of this matter if they can’t produce IN VOLUME? TBD of course..

The Germans may know how to build an ICEV in volumes, their attempts at BEV show they have not yet figured out how to do it.

They still think they need 4-5 years from initial design to start of mass production.
Tesla’s timeline is still to do it in less than two year.

And the Germans have not been able to produce BEVs at acceptable pricing. They are at least 5 years behind Tesla on the learning curve.

the real problem with BEVs is lack of volume. automobile companies make products with the intention of realizing a profit, and it is currently difficult to do that with a BEV. the reason why automobile companies talk in time frames of a decade is not that they can’t make a BEV sooner than that; it’s that they aren’t convinced that the market is there.

It’s easy to convince yourself there is not a market for something if you are not able to aquiesce and make that something.

GM introduced the chevrolet volt and sales were far short of predictions. it’s not the case that GM was not trying to sell the volt, the problem was that they were trying to sell it to the general public and the market is not there yet among the general public. on the other hand, a PHEV is not going to have as much appeal to most EV enthusiasts who prefer BEVs.

The problem, as Nissan and BMW can attest, is that BEVs don’t have a whole lot of appeal, either.

There is a persistent idea that once the right BEV is released, the floodgates will open and BEVs will purchased by everyone; even people who don’t have garages (or even driveways) to park in, people who can’t charge at work, etc. There will be a flood and it will all happen at once.

Unfortunately, we can see in the comments of articles like these that even EV enthusiasts are lining up reasons why the Bolt will not be The One… reasons like:

– I won’t buy from a stealership
– GM isn’t shipping Bolts to Arkansas on day 1, therefore this is just another compliance car
– I can’t drive from Milwaukee to New Orleans in a Bolt

There’s a laundry list of excuses, and there will always be new ones. Maybe the Model III will be sufficiently pure to get these advocates to put their money where their mouth is… but I suspect that when Tesla announces the price for a well-equipped MIII, it’ll be a lot less affordable and the Model Y will be the next magic unicorn that will change everything.

Good points, but market is what you make it to be. If it’s purely based on logic, Pet rock wouldn’t have sold millions and diamonds won’t be worth as much as they do today. Same applies to Tesla (great demand) vs GM (just another boring car company).

As for EV in general, if they find a new source of energy cheaper than fossil fuel, which most likely will be electric of sorts, EV will dominate. It’s simple matter of economics when the flood gates will open.

Great article with some good references. I think Tesla will have produced enough M3s to outsell the Bolt by mid 2018, purely because of the restricted sales locations that GM seems to have imposed post election.

Sadly, as a UK reservation holder, I am resigned to holding on to my polluting diesel BMW until late 2019/early 2020 as I just can’t see Tesla ramping up quickly enough to support all of the left hand drive markets, let alone right hand drive, by the end of 2018.

I always wonder why people think they can the future better with less data.
Tesla has published its expected timeline of development and production goals for the Model 3 a number of times.
There are no factual observations from third party observers that indicate there are serious problems with these timelines. And a number of analysts, journalists and investors did visit both Freemont and the GF to check on the progress.

Then why are so many journalists and commenters, even on a fanboy site like this, convinced Tesla is going to miss its goal by a very wide margin.

Elon Musk has said that Tesla is sold out on the Model 3 until summer 2018. If you have a reservation and all the independent observers are not deceived on a large scale, you can expect your Model 3 before autumn 2018.
Unexpected circumstances not playing a larger role than usual.

maybe people are thinking that tesla will miss it’s stated timeline because people realize that elon musk operates on “silicon valley” scheduling. by that, i mean that schedules tend to revolve around unrealistic “stretch” goals, which are used to push people to meet unrealistic schedules in the hope that in the end they will complete the task more quickly than they would with less aggressive scheduling.

Exactly. Also right hand drive markets seemed to be last in the queue, but no solid date given in the slightest. What else can we do in UK, Australia etc but speculate based on how long it took to get the Model S and X over to us following the US release?

While this article certainly contains some interesting info and comparisons, the start dates assigned to the various models seem rather arbitrary. For example, when an auto maker announces a new model has much to do with marketing, and not necessarily much to do with engineering. Using the date that an article was created on Wikipedia seems even more arbitrary. So, I don’t regard the start dates given here as being authoritative. Also, I think the article could have made this point much more succinctly: That GM does extended testing of early production units before they start selling them to the public, and Tesla does not. As a result, new Tesla models have lots of problems which could be avoided with a more extended period of testing. This is often characterized as Tesla using early buyers as “beta testers”. There is certainly some truth to that criticism. Of course, it’s up to Tesla to decide what works best for them. But I hope that as Tesla matures as a company, it will move to a more extended period of testing, to avoid many or most of the headaches with early production problems. As to whether or not the Model ≡ will… Read more »

This is a great article. Very thoroughly researched and well written.

Since 2016 is almost over, with no sightings of beta cars and no announcement of cell production underway at the gigafactory, I bet that it will be the end of 2017 before the first few Model 3’s are delivered to Tesla and SpaceX employees.


We also know that Tesla has held back on investing in Q2 and Q3, possibly to finalize Solar City acquisition, or simply because the Model 3 design was not really pencil down. Therefore the idea that they can deliver the car within 12 months from start of major procurement, capacity expansion and even battery production seems unrealistic.

I don’t do this regularly. Thank you for your post!
It is exceptionally composed and makes some excellent points.

The big takeaway (now that we are 7+ months into Model 3 production) is that GM STARTS by making 360 cars a week; 9 per hour. Tesla started making maybe 1-2 per hour. In the big exponential growth curve, those scaling from 1-2 up to 9 is a huge time waster (1+ months).