Last 24 hours, Gigafactory sustained rate 3k packs per day, sorry per week, reached peak hour extrapolated outward of over 5k per week, if achieve even once an hour, cont'd refinement, you can achieve sustained rate. What's interesting at least in case of pack production, can do w/ minimal capex, I think in gen, our understanding of production is improving dramatically, exponentially in fact, seek ways to achieve vol w/ dramatically less capex by simplifying prod line, engaging w/ our associates, improving the way parts are made, amazing how everyone has good ideas, just need to solicit and implement them, make ongoing design improvements, scale w/ that if not good in manufacturing then introduce that into flow.
What I've found some things very well suited to manual operation and some things suited to automated operation, 2 should not be confused, I should be clear vast majority of Tesla prod sys is automated, but as I mentioned in tweet a few mo.s ago, we did go too far in automating some silly things, one example we had a sort of, ironically, foolish, we had these fiberglass mats on top of battery pack, basically fluff, we tried to automatic bonding and placement of fluff, ridiculous, incredibly difficult to make work, machines not good to pick up fluff hands humans way better to do that. Super complicated machine, use vision system put fluff on battery pack, asked do we actually need that, realized a part that was unnecessary, didn’t change noise level in cabin, fluffer bot kept breaking down, that’s one of the silliest things I’ve found. We’re also, overgeneralizing the design, so for example, car battery pack has a port for front drive unit, we put sealed blanking plate on, punch hole on it, put blanking plate on it w/ rear drive car, added cost manufact step and failure mode for something unnecessary, that is example of some things, ugh, that’s changed, another result, rapid improvement in batt prod, 7 hours per pack to under 17 minutes now, really radical improvements are possible, improvement in module prod, fully automated now so, we’re able to also achieve sustained rate of 3k vehicles a week, we’re actually slightly ahead in batt module and pack prod than expected, and w/ some work at Fremont veh plant, primarily in gen assembly area, will soon exceed 3k mark, we’re already there in body shop, which also was highly automated, welded the body, already capable 3k cars/wk, gen assembly, w/ some improvement, which will include, temp reduction in automation in a few places, we should reach 3k, basically I feel really good about it Tesla prod in Model 3, really proud of team has done, sacrifice, really intelligent people to achieve this outcome.
Worth noting, see chart in Model 3 market share vs competitors midsize premium sedans, we are almost the best selling sedan in the US in this category, as of April, we will cert be there in May, unless something really odd, then really be there later this year, in Q3, there’s a good chance, Model 3 gets cook-, close to majority in midsize sedans, 30-40% majority market share later this year, this is coming from standing start against a lot of estab brands, very encouraging. Um OK, mmm.
Oh and, yeah, as letter says, I’m quite confident achieving GAAP income cash flow positive in Q3, not certainty, but it does appear quite likely in my view, we are going to conduct sort of a reorganization restructuring of company this month to make sure we’re well set up to achieve that goal, in particular 3rd party contracting companies really got out of control, barnacles scrape, barnacles on barnacles, a lot of barnacle removal, uh alright.
First question? (Elon takes a drink). First question, Brian Johnson, Barclays. Wanna talk about, first of all, production ramp, 3k, 5k/wk run rate, assuming 7/24 or when to get to 5 day 2 shift operation?
Elon: This is what I think, 5 day 2 shift op is ridiculous operation, waste of capex, nor is it the way we oper at Tesla, cell and module production has always op on 24/7 basis, and exception has been general assembly, which is upgraded to 2-3 shifts, 5-6 day 20 hour shift, and paint, upgraded 6 day basis, so I think makes sense operate company on same basis, but majority of Tesla operation has been on 24/7 basis. JB: Aligning everything on same shift schedule makes sense, no inventory build up on diff shops. Elon: Improve capital efficiency is reduce inventory, if shifts not aligned ,build up storage, foolish operate 5 day 2 shift, business sort of we’re using chip out approach, short on short cool --- not like one person working 24/7, it reflects 4 or 5 shifts. Brian: If I do math, tac time of 2 minutes, go back to prior convos, my understanding best in class is 50 sec to 1 minute, grandma walker is 2 times tac time of other factories? Elon: Number, gen assembly not other stuff. JB: You not account OBE, actual uptime of line, make tac time a little faster. Elon: Take a minute, not over 24/7, like sure we did pack prod, today was 32 packs/hour, so we’re under 2 minutes a pack and rising from there. JB: Numbers go up as we go to subassemblies, 4 packs per module, sub factor of 10, 20, higher than that. Elon: I do believe path to manufacturing is velocity and density, that’s what we’re working on rather than trying spending billions on factory, one has to double capex or double prod, or the other is minor capex and eat up the line by double. Brian: I get what you’re saying, some start stop to get? Elon: You can’t have 0 maint time, have to do equip upgrades, ongoing maint, can’t just have operate 24/7.
Next question? Rod w/ Deutsche Bank. Want to follow along on that line of question, extent you’re adding humans in cert automated processes, interpret how these changes affect Model 3 economics, if you’ve done competitive analysis, all these efforts, Tesla prod sys, how competitive vs otr OEMs re labor hours or depreciation/vehicle?
Elon: I’ll say something, unless you have other, the thing I’ve noticed is if you have really complicated machine like buff bot talked about earlier, in order keep it operating have ton of maint and engineering, have to have maint and engineer to fix it basically 24/7, the cost of maint may not be incorporated directly into gross margins but goes into cost of fluff on battery back, I think unnecessary, actually see most likely costs will decrease, cost of prod vehicles will decrease by getting rid of prod stations that are poorly suited to robotics b/c really expensive costs of robot
Deepak: We are really capex efficient let me start w/ that, if look at depreciation cost at steady 5k run rate cars/wk, we’re below most of our competitors, way below $2k/car depreciation basis, as indicated in letter, add’l labor we added but it’s temporary, a lot labor will come out once we figure out smart ways of automating.
Secondly, comments in letter ,advance in battery is interesting, insight how translate that cost/kwh, metric, gains making? Deepak: Every data point we look at internally is best in class. Elon: Best in class of one, we’re the best. JB: Directionally, it’s helpful to understand diff commodities, trends we’re pursuing, reducing cobalt usage is something we’re working on for sev years, very helpful for cost/kwh, commodity price improvement, can’t be quantitative but directionally good trend. Elon: Think we can get cobalt to almost nothing.
Next question? Adam Jonas, Morgan Stanley. Repeatedly said in recent weeks, no need issue equity capital, a lot will say need raise, may not need but will want? Elon: No, I specifically not want to. Follow up: cars produce a large amount of data, SpaceX gets into Elon: Hehehe, not next year, prob 3 years. OK 3 yrs thank you for that, some argue SpaceX can offer secure cybersecurity pipe --- competitive advantage, bandwidth obvious cooperation between Tesla and SpaceX? Elon: could be, cell networks, wifi, still be connected to LEO internet constellation, I haven’t thought about it but probably.
Next question? David Tamburrino, Goldman Sachs. Elon talked about Mod 3 downtime, 1 already happ, otr happ later in quarte,r what addressed at Fremont and what will address little later, are those lone bottlenecks to get to 5k w/in 3 mos? Elon: Tesla prod sys is the best (whew?!) we literally have the 2 biggest factories on earth between giga and Fremont, giga is still slightly smaller than Fremont, but will soon be bigger, Fremont is second biggest building, full answer is complex one, I feel very confident about our abil to get to 5k very soon sustained rate, at Giga, essentially getting to 5k battery packs and motors and power inverters chargers and everything by end of next mo, and body prod, no problem, general assembly is prob our biggest risk, and I’m refocusing on that a lot in the coming month, and then our paint shop is maybe second biggest risk after gen assembly, these are all quite manageable, there’s not like acute brain surgery to get these right, it’s a lot of work just a lot of time and hard work, it’s very doable, and yeah, just really quite straightforward not fundamental impediment.
Software in the car, controls automation, central systems, not even hardware upgrades substantial increases in (velocity?). Elon: John makes good point here, really great prod sys is primarily a software problem, there’s no one in auto indus as remotely good as Tesla, Tesla’s way better at software than any other tech company, the biggest challenge in prod system is software we’re in a good position.
What’s timeline for Model Y, have you spend on this, or hit in 2019? Elon: Not signif until 2019. All the capex this year is Fremont Model 3 Gigafactory? No, please take it literally, it’s not 0 right now but not a big number, relative to – Deepak: only development, not capex, (inaudible) Elon: Amount of money spent at beginning is quite low, at beginning of development program, massive implications for future capex, important to make right design decisions and really think through producability of prod rather than race ahead in capex prod, no question we could have made Mod 3 easier, I’d say Mod Y will be manufacturing revolution, I think will be incredible from manufacturing standpoint, don’t want to go through this again yeah.
Romit Shah, Nomoura: Clarify gross margin comments relate to Mod 3, in letter, said a couple things, over medium term, Mod 3 gross margins will be below target of 25%, also said in Q3, Q4, those gross margs will be highly positive, try to understand what’s poss for Mod 3 gross margs close to 20% by end of year? Deepak: What I say, progressively ea qtr we will get better, and the answer, yeah. Elon giggles. Deepak: It comes down to commodities, currencies, also labor, don’t wanna give specific number but close to it. Elon: Very very close to 20%, could be lsightly below, slightly above. Elon: Sorry furthermore, very confident of 25% gross margin next year def what we expect. Medium term meaning 2018? Exactly, what these things mean, Q4 is we expect to on or about 20%, then middle of next year 25% gross marg, should be where we are, high 20s by next year.
OK, follow up, comment on Jim Keller’s departure, highly respected chip architect, what says about dev of Tesla’s custom silicon and Autopilot? Elon: Jim’s great guy, sort of career he want to pursue for a long time, kinda redesign how architecture works, what I find interesting and ??? Jim I can’t tell, the Tesla, the design of the Tesla hardware is primarily by Pete Bannon, I want to be especially clear, lead designer is Pete Bannon still here, and Karpathy who’s head of AI team, we don’t have plans to hire replacement for Jim’s position.
Next question? Tony Sacconachi, Bernstein. Follow up, prev question gross marg, said once get to 5k units think can get to 25% gross marg on Mod 3, at least 6-9 mo delay relative to what you thought a qtr ago, I’m trying to understand, what key drivers are, labor for cap substitution, don’t think currency sequentially has changed much relative to when you made those statements, help us understand what’s changed for gross marg ramp for Mod 3 relative to before?
Deepak: All in lines of what we said in letter, look at recently imposed tariffs, commodity price increases, weaker dollar, significant material costs, temporarily we use more labor, combine those two that’s what led to our guidance, certainly labor will be address. Elon: Everyon’s talking about 3-5%, that’s something we’ll solve in 3-6 mos, don’t make a case out of it.
Separately, what if anything taking out re lowered capex projection this year, specifically in spending less than $3B, where does that take you in terms of battery and prod cap for Mod 3? Deepak: We ust been much more smarter, like Elon said, we’re not just spending money on automation, looking at problems and fixing?, also being critical about how we grow infrastructure, line it up w/ growth in business, we feel in line w/ decisions and room and discretion in spending money where need. Where specifically? [More?] QUESTIONS NOT COOL, NEXT.
??? RBC cap markets. First question, Mod 3 reservations, some impact news has had, of reservations that opened and configure, what percentage take step to configure? We’re going to go to YouTube, sorry these questions are so dry, killing me!
Galileo, Hyperchange, great quarter, retail investors, autonomous taxies, update Tesla network given Waymo? Elon: Sure, thank you for interesting question, things going towards, shared electric autonomy model, so (eheheh), for whole system to work, need all sys in place, need full autonomy, level 4, 5 whatever you wanna call it, a lot of cars on the road, both software and structure to enable shared autonomy, enable people to share and offer their cars effectively robo-lift or robo-Uber, combo of Uber Lyft and Airbnb, can own your car, autonomous car, available generally to someone to use it, if not use it, recall at will, only use (blabs on about old news). In order to do this we need full autonomy, we’re making really good progress on this, vehicles we’re currently producing, capable of full autonomy, only thing that would really be, might be needed, maybe needed is a computer upgrade, more power to neural net, but that can be done quite easily, really well positioned and building right foundation for you know millions, tens of millions of shared autonomous vehicles, you can not share it if you don’t want to.
Specifically on timing, any details as to when we can learn timing of this service? Well, the hardest thing about timing is regulatory approval, the thing that’s tricky w/ autonomous vehicles, autonomy doesn’t reduce accident rate or fatality rate to 0, it improves it substantially, but the reality is that even though autonomy, even car autonomy reduces prob of death by 50% would be incredible, broadly there’s over a million, 2 mil auto deaths per year, how many you read about basically none of them, but if it’s autonomous situation it’s headline news, and the media fails to mention actually they shouldn’t be writing this story, but how autonomous vehicles are safe, they write fundamentally misleading inflammatory headines so you click on, outrageous, if instead of 1 M deaths you have 100k deaths, people will still sue and say hey you’re responsible for death here, well here 90% of people didn’t die aren’t suing, they’re still alive they just don’t know it, we have to deal with that, and regulators respond to pub pressure and press, if press is hounding regulators, and public is laboring on misunderstanding autonomy less safe b/c press, I find challenge very difficult, and um, yeah, really irresponsible of any journalist of integrity to write article to believe Tesla autonomy is less safe, b/c people might turn it off and die, anyway, I’m really upset by this. Technical standpoint, prob ready by end of next year.
Production capacity, speed on Fremont line, last quarter, max cap was 700k at Fremont, that’s S 3 and X, recent report Reuters, Nov 2019 Model Y start at Fremont, Semi and Model Y where assemble? Reuters report based on nothing, don’t know where that’s from, we won’t produce Mod Y by end of next yr, prob 24 mos from now, 2020 more likely prospect for Model Y, and appx location for Mod Y is not decided, really crowded at Fremont, don’t know where put Y production, we can’t fit it jammed to gills here, one thing for sure is not here, it is crazy packed yeah, we’ll try to figure out optimal location but it’s not here in Fremont.
Not expert in battery back but a lot speculating specs for Semi, even Daimler breaks law of physics- Elon: Haha they don’t know about physics, I actually studied physics in college. Is that linear improvement in battery tech or improvement semi/roadster? Even if not improve batt tech at all we can do 500 mi range truck, we can do. JB: It doesn’t req some dramatic breakthrough, fundamental misunderstanding in what our current prods can do, current status in tech vs otrs in industry, if benchmark best battery they can buy from supplier vs what semi can do, it doesn’t solve, might be where this could be coming from, but basically we have in house can do better. Elon: We can do 500 mi range semi today, actually prod is about 600 mi range.
Also wondering, will go to 350 kwh- Elon: Don’t ask questions that are boring. 350 kw charger, last call JB said sort of will keep status quo w/ your supercharger tech, Elon you said V3 supercharger, will there be one? We will definitely improve supercharger tech, 350 kw charger doesn’t make sense, unless have monster pack or high c rage energy density, it’s kinda cockaminny, I think 200, 350 kw for single car, that’s really pretty, you’re gonna frag battery pack if you do that, you can’t change high energy battery pack at that rate unless it’s very high kwh battery pack, something along the, I dunno JB, 200, 250 maybe. JB: Yeah that’s def power level we discussed and explored, some also go to optimization utility, cost, trade off in car itself, there is trade off charge speed vs range, we look at that carefully, we can design car that can charge faster than 300, 400 kw, but not useful tradeoff. Elon: Energy vs power diff, energy is range, power is peak acceleration basically, you know rate you consume energy, really it’s more important to have long range than it is to have any super fast charge time, you can sorta think about this in devices you use, you rather have cell phone last 2 hours but can charge in 5 mins or 10 mins, but only last 2 hours, or you want something that last 2 days maybe take 1 h to charge. (Thank you, our next-) Elon: we go on while it’s interesting.
Supercharger, would you ever open to other automakers try to generate income that way? We said it’s not intended as walled garden, we will open up, just share cost proportionate to usage, and use adaptor to our chargers, something we’re very open to, but so far no otr carmakers want to do this, it’s not opposition due to us, it’s not walled garden.
Clarify, what’s strategy, sounds like great tactic, build globally, what open it up, not a moat??? I’m wondering why not a mode, I’m curious strategic thinking not a moat? Elon: Moats are lame, if early defense is a moat, you won’t last long, pace of innovation is fundamental determination of competition, if competitors come out something new every 6 yrs, we’re every 2-3 yrs, if our innovation is say twice that of any given competitor, this is true of generally of any industry, any company highest rate of innovation unless killed by nefarious competitor or shoot themselves, will exceed competitor, case w/ Walmart vs Amazon, outcome was obvious a long time ago.
Megacharger, will sell energy at fixed rate for operators, operate at cost and reduce rate, or use as energy stream? JB: We haven’t talked or finalized any of that, we want to make sure seamless or easy system for trucks, some customers may elect to work w/ us whole system or parts of it, a lot of ways can be solved. Elon: Commercial trucking, heavy duty semi, can always look fundamental, not deciding based on aesthetics or consumer related things, we try make our Semi cool and sexy b/c we think it’s good to do not b/c it aff buying decision of customers, we have laughable lawsuit from company ironically called Nikola suing Tesla, hehehe, Nikola suing Tesla, that’s absurd nobody buys b/c how it looks wraparound windshield, we have to make sure Megachargers Superchargers whatever they’re called set up so trucks have very low cost to energy. JB: Super exciting, potential to link up to renewable energy generation to future trucking fees, give customers incredibly determinative costs won’t change w/ cost of petroleum price in decades we’re excited about. Elon: Emphasize, trucking companies, if cost of diesel goes up a few cents it destroys their business, we’re with them sort of megacharger situation, combining solar, battery, Megacharger, costing costs we know what they are, it’s very predictable, and lower cost per mile vs diesel truck, fundamental is cost per mile kilometer of cargo that drives the commercial trucking market, can have ??? truck in world.
How trucking market can change or grow if you can do things w/ dramatic cost reduction like platooning? Elon: Take revenue from railways, rail cars is platooning, small crew operate trains, trains don’t go everywhere so have trucks deliver things to train rail spur, then pick up from rail spur there, truck plus transfer plus rail spur, platooning will dramatically affect rail industry in negative way.
Tesla energy, I assume supply not demand constrained, how prioritizing residential vs ???, how S Australia change perception in battery storage? Elon: Quite profound effect, S Australia quite well, far beyond their expectations, battery can respond at ??? level far faster than any hydrocarbon plant, value saving proposition even better than gas turbine plant, get on Tesla instant acceleration, like mind melt w/ car, the car is you, the same rapid response is true of battery pack, utilities work w/ us really love battery pack, I feel confident we’ll be able to announce a deal at gigawatt hour scale w/in a matter of months, a thousand megawatt hours. JB: First part of your question also, absolutely accurate, more than enough demand, still building out of demand backlog, increasing it slightly, trying to do our best to prioritize customers between residential commercial utility, longer term strategy is to shift a little our focus catch up on powerwall demand backlog, it’s too long right now people are waiting too long, generally direction we try to take that, Mod 3 take a lot of focus in last few qtrs., trend will reverse in second half of this year.
Next question? Phi Labeau, CNBC. Tesla Semi, perspective, how many reservations you have now, where you are in the plan for dev it, rolling out first model? Elon: We’re discussing something, sorry, repeat question? (Question repeated) I don’t know how many reservations we have for semi, not 2k, if we really try Semi, it’s not like ongoing sales efforts, orders for Semi are like opportunistic, companies approaching us, our focus is Mod 3, we need to get it above 5k/wk, good margin, become profitable company, good and accurate criticism, high time we became profitable, truth is you’re not real company until you are frankly, that’s our focus right now, we got an awesome product roadmap, Tesla Semi is one of those things, I have a really good idea, Mod Y is going to be amazing, really excited about that, Tesla Pickup’s gonna be great, so uh, prod roadmap we have way more cool things than we know what to do, idea generation we need to execute it, we need to not divide our attn. to too many things at one time.
When will oyu make decision re second manufacturing facil? Elon: That’s prob later this year, has to be later this year, maybe next qtr, but not later than Q4 for Model Y, and we also expect to announce Tesla Gigafactory in China soon. In N America or China? I just said in China. Gigafactor, second manufacturing plant in China also? In future all gigafactories will include vehicle prod, right now battery prod, motor will be in gigafactory, future will incorporate all production. We’re appreciative that government of China will allow full ownership of manufacturing facilitiy, word of appreciation to Chinese government, Rob anything you want to say? We are in good discussion w/ government. Elon: So Robin Rens is here w/ me, managing world wide sales for Tesla (inaudible) um I think next CC, we’ll have a lot more to say about that in the future.
Next question, James Aberteen, consumer edge. Briefly, given coverage you’ve received, high profile accidents, one thing we like the best about your company, test daily from Autopilot perspective, data re confidence re Autopilot function, use more or less, purchase func more or less in lieu of accidents, get sense of consumer, likelihood consumers adopt tech over time? We do see steady increase % of miles driven as we make Autopilot better, see steady increase, I think something on order of a third hwy miles, maybe ½ in some regions on Autopilot, of course if negative news in press that dips, OK that’s not good b/c people are reading things that use Autopilot less that’s not cool that gets me upset, we get accused of blaming victims, it’s important people not be under impression, statistics unequivocal, Autopilot increases safety, we will also publish safety statistics on qtrly basis, ppl will know exactly what Autopilot safety is, getting better, worse, common impressions, if accident on Autopilot, people for some reason think, driver thought fully autonomous and wasn’t and we somehow misled them, it’s opposite case, when serious accident, almost always, or always, it’s experienced user, and issue is more one of complacency, get too used to it, that seems to be more of issue, not lack of understanding of what Autopilot can do, more issue, they think they know more about Autopilot than they do, quite significantly.
Two accidents, had dips during those periods, you are seeing increasing usage and weathering those dips based on where we are today? Correct.
Next question, Ben Kallo, Baird. I remember the Bear? Story, hung up on battery costs, but stakeholders shareholders, so far couple push outs, can update us on 3k or 4k reach numbers, you’re active on Twitter, a lot of news out there, volatility, hard to own stock, even though myopic I think important to give those updates, give us updates when get to 3k or 4k/wk on Mod 3? Elon: Yeah Tesla is such a leaky server of info, it’ll leak easily, and people track reservation regularly, any info we provide will be only a week or two ahead, people track registration and shipments, problem is people focus on what’s happ in space of weeks or months, but maxim of investing is focus on long term rather than short term, we have no interest in satisfying day traders, sell our stock don’t buy it. I completely understand, great quarters, years, volatility has way of shaking people out, drawn out, anything can do to help? I think if people concerned about volatility shouldn’t buy our stock, I’m not here to convince people to buy our stock, don’t buy if volatility is your concern.
Last question? Alex, Piper Jeffrey: Not sure if hard question to answer, mentioned Mod 3 market share vs 3 series otrs in that segment, what extent will Mod 3 change denominator make that segment larger as class vs what it used to be? Elon: I think will increase total number of sedans, I think so. You’re thinking polling X accord and camry buyers into that class, as example? Yes b/c trade ins, we’re seeing quite a range, not necessarily Audi A4, 3 series. Deepak: Also Model S, Model 3 more prominent. Elon: Yeah, also once we get to shared autonomy, ridehailing, could be as soon as next year, it’s technically ready, expect some jurisdictions will give regulatory approval, cost of ownership drop b/c can share w/ otrs.
Last question, Model Y will be true manuf revolution, for Model 3 if can do over, what specifically would you do or plan to do? Elon: Save that for another time, we’ll talk about that when we unveil the Mod Y, it’s really gonna be dramatically better, the design and prod sys will be really next level.
Thanks everyone for your questions.
Disclosure: I am/we are short tsla.