I used to work in a stamping and weld plant, adjusting stuff to hold tolerance. I remember being blown away by the giga castings and how much pain they avoid. then I was amazed by the floor battery. this is gonna be great. oh and BTW the buffering thing is not really a thing; even if you keep the rest of the line moving, storing parts in inventory, once you fix the blockage it's not like that station can suddenly work faster to clear the buffer. Toyota figured this out long ago and everyone has basically followed suit.
Tesla is able to build faster because every factory 🏭 uses , warehouse on wheels, and AI factory 🏭 software to schedule parts as needed. no more bulky & expensive warehouse.
when you say it's not a thing, do you mean it's not done or just that it doesn't help with speed? I can see the buffer being held so they can be used if that sub-assembly goes down in the future, but it's also understandable it might not be worth the floor space to do that for all sub-assemblies.
@@qfrank76 its generally not done because it has been shown to be inefficient. Toyota has all sorts of materials referencing "pull manufacturing" or "kanban" if you wanna read into it further.
I imagine the Tesla employees as curious folks who explore possibilities of further automation even if that means their job is replaced by a bot, because they're moving to another job as it goes. The contrast is the unionized dock workers, whose jobs are already replaced by technology (saw some fully automated port operations in China), yet they threaten to hold the entire country hostage if their job should be done more efficiently by the automation.
"Knowledge mitigates risk. Wherever you choose to invest, become an expert in that field." You've always provided the best and most comprehensive knowledge for Tesla investors. Been here since the first video. Thank you!
Great video, thank you. Unboxing also enables the integration of humanoid robots like Optimus with much less complexity. A robot working on one sub assembly that doesn’t require “craning” its neck to get into the vehicle is going to massively increase operational efficiency. And you only need to “bite off” one part of the vehicle at a time. Tesla pays $25/hr or so ($32.50 fully loaded) per person on the line. Optimus is estimated to come in at $1/hr. Even moving to 10% Optimus installed on the sub assemblies will have a massive opex impact.
Manufacturing engineer here. There's a little more to this and I would be happy to have a little discussion if you're interested. You would be even better off finding a staff automotive manufacturing engineer though, because there will be even more stuff that they know in specific (capitalization tradeoffs, specific unit manufacturing issues that unboxing solves, etc) that I might know only in general
engineers don't have a clue. It was the "engineers" who failed to consider human body-torquing where autoworkers needlessly contort, a scheme guaranteed to produce vehicles that will be recalled with massive warranty costs - for decades...
Thanks Jordan! One advantage you didn't mention is foreign factories. For Mexico/India/Indonesia, they could ship completed modules (volumetrically efficient, as you point out) and just do final assembly at those locations. They can start with a small factory, and expand upstream to locally produce modules as production scales. Or it could also allow for micro-factories that only do final assembly where there is low volume for the local market, like Spain/Italy/Australia/Japan/Argentina, with gigafactories feeding completed modules.
Thanks man! That could happen in the deep future when these things become commoditized, however... Tesla does all mission critical work in-house They only outsource non-mission critical items Major vehicle components are mission critical
TESLA doesn't have any assembly only plants anymore, it closed Netherlands many years ago. Gigaberlin is hub for EU except RHD. China is main exporter of RHD and those markets. Fremont & TX are exporter to Canada & MX. for Cybertruck & some other vehicles, like NEW model 3.
@@thelimitingfactor Thanks Jordan! Sorry I said "modules" but I really meant "major unboxed components" like front-of-car, SBP, etc. They would come complete from a Tesla gigafactory, not outsourced. And the local micro-factory would just bolt the few pieces together in a final process.
@@thelimitingfactor Currently shipping cars is volume intensive - there is a lot of empty space being moved. Relatively inexpensive mini factories where the final assembly could be done without any of the background requirements (casting or paint shop for example) could enable a great deal of flexibility. The components could be shipped from any Giga with spare capacity.
SUPER! Thank you for going into the details of the inneficiencies of the traditional assembly processes, and all the ways that the unboxed process addresses them.
this presents interesting opportunity's or trouble for collision repair. If the car is 6 major components, then the damaged sections can simply be replaced. Interesting to see how this is implemented, I see a lot of structural adhesives and SPR's being deployed
We all missed one important early step, executed years ago! Early Model S had an opening sunroof, with all the trim, mechanical complexity, sealing, and lots of weight. Long ago, Tesla scratched all that and began to build all models except for X with structural glass roofs, which are simply glued on! Piece of glass and glue. Can't get much simpler than that!
@Withnail1969 sure, it's a feature many want; no argument there. However, it's a genius leap in simplicity to go with fixed glass; easier than mere sheet metal. In the context of this video, that's what's relevant.
@@Withnail1969 - #1 SAFETY, #2 it's a structural component of whole car, Tesla roof is Stronger than stamped roof, Tesla roof broke a car crusher, trying to recycle. # 3 speed of production & reduction of multiple process. and MORE.
The design of the vehicle is key to improving the build process. Many vehicle designs would be impossible using unboxed but Tesla makes EVs only with very clever structure designs so they can adjust everything as needed to get to the next level of optimization. Other automakers can't do this because they rely on parts vendors for some of the car and don't have control over it's design to some degree.
Fine video but it's rather more complicated than you think , years ago I was given a job to stop problems with trim in the assembly of the Vauxhall Frontera SUV which was being assembled from mechanical kits from Isuzu , the issue was failure to adhere the plastics due to the temperature of the vehicles coming from the "colour buffers" these were cold and were the strategic reserve for the production process , the issue caused rejects and line halts at assembly and inspection , I designed a IR heater to monitor the shell temperatures incoming onto the assembly and adjusted the heating required , it worked perfectly and allowed mixed colours of shells to lifted to the correct temperature on the fly , the shells varied by 30 degrees F depending if they were a freshly painted shell or were from colour buffer so could be days or weeks old , the colour line in the paint shop ran in batches depending on popularity and colour buffer space , there are thousands of points in the assembly that the smallest thing can cause a line delay , most of which you would think would be trivial but the consequences are enormous later down the line
Another thing that stikes me about the unboxed process is there is no need to have the same number of lines for each portion of the vehicle e.g. Takes twice as long to construct the front of the car as the doors? Just have two parallel linnes for the front of the car and 1 for the doors.
I would appreciate a video about all the reasons other manufacturers haven't used an unboxed process and the ways those reasons either don't apply to Tesla or have been mitigated by Tesla.
I can give a few points, imo at least. Revamping factories is a lengthy, expensive process that requires a lot of capital and expertise. Many OEMs do not have shareholders backing them for these sorts of investments, they prefer share buybacks etc. It has also never been done to this extent so trying something new is "dangerous" to their existing business. They usually prefer the "wait and see" approach once they have reached a certain stage in their business. Also, the other companies don't have someone like Elon willing to double down when he feels the need. Plus many other factors, if the above helps I can continue, let me know. Enjoy your day. Trying to help you out if I can Jordan 😊
@@jessstrap2088 The problems expand when you then factor in a workforce that's also trained in processes that will now need to be changed or scrapped entirely. That means re-training or pensions and in a UAW environment neither is cheap.
Automation is simplified and easier to update with larger subassemblies. Assemblies for different models may be produced on the same machinery. Castings may be batched for example by changing the molds only.
Great video. You fill in many of the gaps from the presentation. I only wish Tesla would use this manufacturing expertise & excellence to build cars like the XPeng MPV 9 or the Li Mega. Both have amazing aerodynamic drag for their size, more seats, middle row massage seats, a minifridge, and a rearview mirror that can switch to a screen for the rear camera. Everything people want in an EV limo or family hauler. The best thing Cybertruck has done IMO is make cars like those MPV I mentioned look sleeker and more normal. Now if only I could get the $55k XPeng in America for that price!
Thanks! Teslas tend to be more utilitarian except for the very high-end vehicles They've found that a lot of people don't use the bells and whistles Most people would rather have a lower cost vehicle
@@FrunkensteinVonZipperneck...so Elon took on a difficult Cybertruck project while Tesla only has 2 sedan and 2 SUV (1 affordable of each) for sale. Think back to the relative ease of converting Model 3 to Y. Tesla could've built a light duty minitruck on the Y platform, assembled on the (too-often idle) Texas Y line while they tweaked & perfected Cybertruck. A small truck ready for international homologation. Instead we have an overpriced overweight Cybertruck, the ire of EU & Asian regulators, which lacks features of similar-priced Chinese MPV. Fans & shareholders can still be critics. *Should* be critics, imo.
Jordan - Tesla in 2024 a partial unbox process, and likely use Advance UBP in near future, as it works out Full UBP. and Tesla will have many iterations of UBP in future.
In the unboxed method buffering should actually be required in sub assembly areas so if there is a slow down the extra parts made in advance can fill the gap while the problem is solved. Over time, common slowdown issues could be identified and buffer size could be adjusted to only be enough for the worst case scenario. Buffering can make the best use of vertical space like and upsize down pez dispenser.
Thanks Jordan, completely Professional as usual. The OEM’S are in real trouble and I am sure they know it. But, what to do when you have so many factories and so many staff and so much debt. Where to start? My suggestion is the Boardroom and they only get the same as their workers severance payout.
How will all the components be joined? Bolts, glues, welds, tab & slot, special techniques for mating castings (aluminum, maybe magnesium) and stamped parts (steel)? Guessing continuation of Tesla's experience joining structural battery pack & castings. It might be fun to learn specifics.
The Gigacasting are directly joined to structural battery pack w/ bolts and adhesive. Tesla has perfected wielding Gigacasting to stamped steel & extrusion. but is constantly switching over to partial UBP. modules are becoming high quantity of cast parts , rather than stamped steel, extrusion, and small cast parts. eliminate brackets and use CRP moulded over extrusion or castings.
Rover cars had a conveyor that ran at one speed only. It would have been difficult to change it. The 400kW infra red sill oven baked the sills based purely on the line speed to get the desired temperature profile, with no electronic temperature control at all. Unboxed process allows different station process speeds throughout the factory.
I just had a thought. Is there any company which has this many, high quality and in depth, youtube channels as there are for Tesla? There is so much going on it's wild! If there are other companies with such a dedicated following I would like to hear their names. I guess Apple and maybe Meta, but it seldom feels like they are trying to solve important problems. (I love the Apple Silicon chips, but there was almost nothing out there for speculation/analysis etc. afaik)
Yes, there's no comparison 🤠 CATL and BYD are doing some cool stuff, but their presentation are always dripping with technobabble and the only really operate in a few domains at most. Tesla's pushing the limits in over a dozen different domains.
In theory the "Unboxed Process" allows for significant redundancy similar to a modern server stack where there is no single point of failure. All the sub assemblies can be duplicated and run parallel; this offers: *redundancy, *expandability & *live-upgradeability. Indeed the "Final-Assembly" can be expanded out to more than one line for the same reasons. (Sidenote: The CT door autopsy showed the modular door subcontroller with only one connector to the body - no loom. All part of the evolution towards this new process).
Tesla is almost only OEM, that uses discreet axle modules , that use high amount of casting, as opposed to wields, fastener, brackets, and different materials, like stamped steel, extrusion, and small cast parts. the entire axle can drop out to be worked on, just a few bolts.
11:27) Jordan, "casting process" is necessarily asynchroous: faster casting is required to store part inventory to be used during required downtime - for switching out dies and maintaining GigaCast machines ...
Great video, Jordan. You did not mention Tesla One, the overall manufacturing management system. Which manage also spot changes and manage necessary relevant changes on other systems. This will be much easier in the unboxed process. It can also ensure that buffering mainly occurs in the inventory section and that the line itself hols the optimum amount of sections needed to complete the part. It may even be that the unboxing proces spreads out to the other parts of assembly line. Like for instance building the structural battery back with additional elements like chairs, suspension etc. Tesla can provide information to see if it is more efficient to have an element move to another part of the subassembly. Exciting times.
Thanks man! Yeah, there's a WHOLE lot I didn't cover here. It could have been a 2 hour video easy. Overall good additive points. Thanks for the comment.
@@thelimitingfactor Yeah, so True. If you decide to do a deepdive on Tesla One, maybe do a co production with Joe Justice. He seems to be on top of this.
About 40 years ago Volvo developed and implemented a similar process as the proposed Tesla unboxing process. Volvo did not continue with it and returned to the traditional production line approach. Tesla has the advantage of gleaning lessons learnt from the Volvo attempt as well as having the advantage of more advanced technology being available today.
These ideas are rarely new. I wonder how franken-car stitched from 10 parts will look and behave in long run. Of course, every car is built from parts, but each part of process allows for alignment to happen at end of that point. I guess consumers will find out in the end.
I suspect the advances in automation and manufacturing precision in the last 40 years will make a big difference towards making the process more viable. something like the unboxed process requires very tight tolerances to go together properly. The gigacastings with integrated brackets instead of lots of separate parts also dramatically reduces compounding errors, so that's probably a big part of what makes the unboxed process even possible even with modern precision and automation.
Amazing explanation as usual. These tech innovations synergize in interesting ways, now I can only imagine that if Optimus were added to reduce humans in the unboxed production lines, optimizations, speed increase and cost reduction could grow even further. Seeing how 10/10 logo says "we robot" I can only hope all these improvements will make it possible to build not only the robotaxis but also the affordable $25k compact car as well. It's not only that people in the world need those to exist for the EV revolution to keep happening, it's also that detractors arguments that ICE cars are still cheaper will be squashed and won't stand anymore, so it must happen, it must be done, the world needs it to happen and everyone who trusts Elon is hoping he can make it happen, as he has shown he is the man that makes the impossible, possible.
Tesla AI factory 🏭 software is underappreciated. every Tesla made , has a digital twin in software, and in 2024 its final assembly is done w/ computer installed, and running in Service mode. this is to reduce mistakes & alignment.
I like the production design and the new technologies for the Cybertruck. And I like the concepts to be employed in the unboxed process. But my hope and question is whether and when these advances will be applied to the Model 3 and Model Y (not just for CT or suggested future vehicles). Maybe that would be a long while; require new factories (or perhaps there is room in some existing factories for new assembly lines); or maybe actually never for these models. The "trouble" (odd to call it that) with EVs is the rate of technology advancement. Like computers, both the factories and the products are almost obsolete when you buy one. About the only thing that can keep track of that (and assist when service is needed) is the software that defines and runs both.
There is one big change we haven't discussed and that is how to fix the different vehicle modules once they are sub assembled. Welding is probably not an option. Using screws is challenging as they have to be many or there will be stress concentrations. I would guess glue plus screws would be the best. On the Cybertruck with huge castings and very strong panels, screws are sufficient.
The Ford Model T was produced with a process fairly close to the unboxed, however that required a separate frame which is a big drawback for the final product.
I always imagined that by having sub-assemblies it enables multiple Vehicle Assembly lines in parallel. Current lines are 30-40s for a car to be completed off the line, but instead have multiple assembly stations (rather than a line) to assemble each car... wont need snaking sequential line running through the factory but rather just nodes or almost pitstop like areas where the assembly is done and drive off?
So Tesla is risking the paint colors not matching as well and either skipping spot welds between modules. In return they get to decouple the design of the 6 major sub-assembly lines.
The traditional reason for painting the fully assembled body instead of individual parts is to produce a more even and uniform paintjob, especially with fancy metallic-effect-paints that produce shiny reflections and patterns that are supposed to appear uniform across the entire car. Depending on available paint options that may be something to think about.
@@thelimitingfactorI think powder coating would be the fastest and most consistent as it is applied dry and the smaller parts could then go through a continuous oven. The oven could be as small as 4’x5’ opening…. Then again no paint would be best.
In 2024 oems are Still having to rework stamped parts with 2x4 wood, and 🔨......lol. Gigacasting requires zero rework or post treatment. Cast parts fit perfectly, everytime.
What I don't understand about the unboxed process is that it is said to take less factory footprint area. It seems to be that the unboxed process is similar to paralell processing in a CPU which utilises more die area, though admittedly here more throughput is achieved via multiple channels whereas with the unboxed process the same amount of assembly is carried out in less time. It seems to me that thought the lengthe of the main assembly line will indeed decreace it's 'width' will increace. Does the unboxed process really take less factory footprint area, the same, or more?
@@thelimitingfactor Thanks for the reply. I still have my doubts that it even creates more output per unit of factory area however. I can see that it reduces the time from start to completion due to improving ergonomics and parrellisation but if you have to provide workstation areas beside the production line this surely must require more space. The length, (and area) of the main line will indeed be reduced however. Be careful of conflating the concepts of rate of throughput with rate per unit of area.
So... my imagination got the better of me and I pictured making hamburgers. Current car assembly lines are like this: The burger bun is baked as one piece, then buttered. It is then sliced in half and separated, and then carried on a conveyor through the length of the building. Meanwhile, the hamburger patties are carried along, just an inch above the bottom bun, and tiny workers climb in the gap and spray condiments on the bottom bun. The top bun is only 2 inches above the meat, and workers use robots to bring in the lettuce, shove in the onion, and try to stuff in the pickles and tomatoes without them sliding off the outside. Then some crazy South African comes along and says, why don't we cook top and bottom bread rather than a bun we have to bake, cool, slice, butter, and carry the length of the factory? We can have 2 ovens, one on each side. Then we bring in the bottom bread with condiments already applied, meat with the veggies already sitting on top, then stack the top bread. Stacked all up, made fresh and more quickly! Think outside the bun, y'all 😂
What's your thoughts on Solid state batteries? Especially since MG claim they will come out next year. And comparing to Tesla's battery strategy. Could make an interesting video?
We can hope, but so far they're a lot like Fusion, always a year or two away. I think Toyota has been planning to produce solid state batteries next year for going on ten years now.
By making the paint shop more efficient, they also lose a major hidden issue. Paint. Automotive paint has to be stored within a tenth of a degree. It’s how a lot of long term warranties on body’s are achieved. Less paint less storage, less energy. Better paint efficiency also means less pollution. VOC’s are a problem for paint shops to control, and those controls are maintenance heavy. Often they use wet booths, however Tesla have already started to attack this point with a dry system. So less water usage.
TOPIC SUGGESTION: If a battery is cooled from the top and the bottom, can you create a battery 1000x80. Like 1 meter diameter 80 mm tall big pizza sauce coin cell battery. It can be used for grid storage only.
@@thelimitingfactor I suggested this for stationary storage where volumetric and gravimetric efficiency don't matter as much. What matters is the ratio of active to inactive/structural material.if you have a stainless steel barrel and inside you have a stack of jelly rolls separated by cooling plates(disks) that also act as current collectors. This approach will improve manufacturing speed and reduce cost
Are you sure? Modern vehicles use adhesives to augment stress concentration areas ( welds, bolts). E-coat oven cured adhesives are higher performing than ambient temperature cured adhesives. Unboxed requires ambient temperature curing.
Thanks for this, nice summary. Not mentioned were steer by wire and brake by wire, implemented in the Cybertruck. This is enabled by the Canbus loop, that you did mention. Removing steering wheel shaft and hydraulic lines for brakes helps the unboxed process by removing difficult linkages between sections of the building. Enabling a plug and play assembly.
What are the cons? Could it be that less parts then it's more expensive to replace each part? If the car gets hit behind the back, it doesn't seem to be an easy fix
When they assemble the finished parts, and they don't align within tolerance, maybe it would be harder to get them into alignment without damaging paint/finishing. I hope they'll be designing the fastenings to be able to handle this instead of the cybertruck "make it all be micron level accurate", and when it isn't you're screwed. Other "cons" is that this is a new/untested process. We don't know how well this will end up working in practice. That goes for any innovation though. Crash worthyness and lifetime may be negatively impacted by this, since the overall structure of the car is somewhat different because of the fastenings. These aren't unsolvable, but new is different.
keep in mind, the WORLD will go Gigacasting, structural battery pack & unbox process. Cherry auto in China, now using 16,000 ton Gigapress to make complete Underbody. and TK press makes 9kT, 12kT, and 16kT presses. the 16k ton gigapress, is same size as apartment building on its side.
The question is can this produce a quality product... its one thing putting together a truck or a bottom of the barrel cheap vehicle, and another thing putting together a high end product where long term comfort and ease of use, and very low fault tolerance, are the norm
It already will be improving quality. Giga casting creating a solid frame for the panels to fit to. The battery and seat sub assembly cutting down on through door mounting damage risks.
@@thelimitingfactor i am wondering if any process in the traditional manufacturing is a bottleneck- if so I assume this kind of „optimization“ would help to increase production speed further
In most large organisations, the dooers and innovators get weighed down or stopped completely by the parasitic departments and functions that spring up around them. Many young people walk unto these companies with fresh ideas and are crushed by the corporate machine that rewards mediocrity and greasy pole climbers more than with long term players trying to help the company.
If you have precise location points, for instance a pin on one part fitting a hole in another, you will get a far more precise and controlled structure with minimal cumulative variation. Then, as Jordan says, you can bolt them together. Add that to the precision of the megacastings and the overall tolerance of the complete vehicle dimensions should be impressive. . Jordan.... Another advantage of bolting together? Major repairs? Let's say there's front end damage, severe enough to require a Casting change (I know.... Unlikely not to be written off) . The whole assembly process is *designed* around simple and rapid connection between sections of the vehicle? As such logic would suggest easy mechanical and electrical *disassembly* of the front casting "unit" from the main body, possibly allowing for that major repair by bolting on a new casting preassembled with old and new subcomponents as required rather than "bending" a conventional body to shape. . This may in fact lead to FEWER vehicles written off?
Outsourcing the one thing they don't want to do. Have you not heard about vertical integration and how outsourcing has delayed and causes problems in the legacy lines? It's worth your time to do some reading if you're interested.
I think this opens up more options for the subassembly lines. One subassembly backing up more than the others? Could put in two subassembly lines for that.
@@thelimitingfactor More likely that if you had let's say the rear cast and battery/ interior section running "fast" you would simple stack the extra units and pause those sections.
why not painting insides? tesla can rust, and any aluminum alloy can be real bad with it. Some people used tesla cars in high humidity countries ended with rusted through areas (as they said - turned in powder). Tesla car shouldn't last for 5-10 years, that'll make it's most expensive part - battery pack - useless (that can last for 30-40 years)
Hmmm, I haven't heard anything about that about the rust. I'm curious which years that was for. As for the battery. No, it won't last 30-40 years. More like 15-20 depending on the chemistry.
@@thelimitingfactor rust goes from wet dirt that that have some salt in it. PArts that involved most in it is places behind wheels (mostly, front). Thin anti-corrosive layer goes down from 5-10k miles(maybe faster), then dirt placed there starts the rust and if that is started, there is almost no way to stop it since it goes deep into metal. Batteries are just guaranteed to use in 15-20 years, they can last a lot longer (and there's no real risk of car self burning if chemistry is LFP) so that cars and used battery packs will be there, it's only matter of time.
One possibility you didn't explore is having redundant sub-assembly lines and some permanent buffering of sub-assemblies. This would allow production to move to the backup while a line was down and keep up the vehicle production while a sub-assembly line was being addressed. It would only make sense if there were multiple lines for each sub-assembly and if stopages were common enough. Less efficient than the designated sub-assembly line in operation, perhaps, but still maybe better than a main line shutdown.
This seems very capital inefficient. Unnecessary redundancy. Creating parallel production lines gets rid of the bottlenecks and makes main assembly simple It atomizes and modularizes everything which reduces issues
@@thelimitingfactor Having a main production line down has an hourly (actually, often by the minute) cost. Wages are often still payed, at least in the short term. Overheads continue unabated and cost of capital is undefrayed. While having "just in case" storage and production is rarely cost efficient, a small amount of redundancy/storage can be the most efficient by avoiding these stoppage costs. Indeed, Toyota found this out quite a while ago and aims for "most efficient working capital", not "lowest possible working capital". The US majors didn't pick up on this, which is why, initially, Toyota fared better during the pandemic. As an example, I heard of one case where a vital part in a car line died and the company flew the replacement in for around $100,000, many times the value of the part. However, the failure was so rare that it didn't pay to have a replacement everywhere the part was used and the cost of a stopped line was so high that the cost of the flight was lower than the cost of the line being down while the part was driven there at far lower expense. The optimal approach is, I suspect, a linear programming problem dependent on the frequency of line failures, the average length of time to fix, the number of identical lines, the capital cost of a redundant line and so on...
unbox process is the future, and AI factory 🏭 software. almost no one else uses , warehouse on wheels in auto manufacturing. and many years ago , Tesla bought Rivera tool & die. and Canadian assets. Tesla can make ANY tool & dies it needs for production.
Perhaps they've gone from being below industry standard with paint to figuring out how to be above standard and more consistent such that parts don't even have to be painted together. One can hope.
@@thelimitingfactorThe paint match is why the doors are mounted first and removed later in a 'boxed' assembly process. Solve that problem and a 'boxed' process can do away with that inefficiency as well. Nothing to do with boxed or unboxed.
Rover cars had a conveyor that ran at one speed only. It would have been difficult to change it. The 400kW infra red sill oven baked the sills based purely on the line speed to get the desired temperature profile, with no electronic temperature control at all. Unboxed process allows different station process speeds throughout the factory.
I used to work in a stamping and weld plant, adjusting stuff to hold tolerance. I remember being blown away by the giga castings and how much pain they avoid. then I was amazed by the floor battery. this is gonna be great. oh and BTW the buffering thing is not really a thing; even if you keep the rest of the line moving, storing parts in inventory, once you fix the blockage it's not like that station can suddenly work faster to clear the buffer. Toyota figured this out long ago and everyone has basically followed suit.
Tesla is able to build faster because every factory 🏭 uses , warehouse on wheels, and AI factory 🏭 software to schedule parts as needed.
no more bulky & expensive warehouse.
when you say it's not a thing, do you mean it's not done or just that it doesn't help with speed? I can see the buffer being held so they can be used if that sub-assembly goes down in the future, but it's also understandable it might not be worth the floor space to do that for all sub-assemblies.
@@qfrank76 its generally not done because it has been shown to be inefficient. Toyota has all sorts of materials referencing "pull manufacturing" or "kanban" if you wanna read into it further.
I imagine the Tesla employees as curious folks who explore possibilities of further automation even if that means their job is replaced by a bot, because they're moving to another job as it goes.
The contrast is the unionized dock workers, whose jobs are already replaced by technology (saw some fully automated port operations in China), yet they threaten to hold the entire country hostage if their job should be done more efficiently by the automation.
"Knowledge mitigates risk. Wherever you choose to invest, become an expert in that field."
You've always provided the best and most comprehensive knowledge for Tesla investors. Been here since the first video. Thank you!
Knowledge is power - schoolhouse rock.
Thanks for the long-term support man! Much appreciated!
Jesus freaks smoke weed?
@@longmuskox4194 This one does.🤣
@@longmuskox4194 Jesus is my Rug man. He really ties the room together.😎
Great video, thank you.
Unboxing also enables the integration of humanoid robots like Optimus with much less complexity. A robot working on one sub assembly that doesn’t require “craning” its neck to get into the vehicle is going to massively increase operational efficiency. And you only need to “bite off” one part of the vehicle at a time.
Tesla pays $25/hr or so ($32.50 fully loaded) per person on the line. Optimus is estimated to come in at $1/hr. Even moving to 10% Optimus installed on the sub assemblies will have a massive opex impact.
Manufacturing engineer here. There's a little more to this and I would be happy to have a little discussion if you're interested. You would be even better off finding a staff automotive manufacturing engineer though, because there will be even more stuff that they know in specific (capitalization tradeoffs, specific unit manufacturing issues that unboxing solves, etc) that I might know only in general
The best way to contact me is on Twitter
At limitingthe
@@thelimitingfactor C-V2X connected cars as a way to improve FSD is something I am surprised to not hear more about. chicken and egg situation?
Can’t wait the hear about this discussion!
engineers don't have a clue. It was the "engineers" who failed to consider human body-torquing where autoworkers needlessly contort, a scheme guaranteed to produce vehicles that will be recalled with massive warranty costs - for decades...
@@FrunkensteinVonZipperneck those were Big Auto engineers. They weren't led by visionaries, but short-sighted bean counters like Bob Lutz.
Thanks Jordan! One advantage you didn't mention is foreign factories. For Mexico/India/Indonesia, they could ship completed modules (volumetrically efficient, as you point out) and just do final assembly at those locations. They can start with a small factory, and expand upstream to locally produce modules as production scales. Or it could also allow for micro-factories that only do final assembly where there is low volume for the local market, like Spain/Italy/Australia/Japan/Argentina, with gigafactories feeding completed modules.
Thanks man!
That could happen in the deep future when these things become commoditized, however...
Tesla does all mission critical work in-house
They only outsource non-mission critical items
Major vehicle components are mission critical
TESLA doesn't have any assembly only plants anymore, it closed Netherlands many years ago.
Gigaberlin is hub for EU except RHD.
China is main exporter of RHD and those markets.
Fremont & TX are exporter to Canada & MX. for Cybertruck & some other vehicles, like NEW model 3.
@@thelimitingfactor Thanks Jordan! Sorry I said "modules" but I really meant "major unboxed components" like front-of-car, SBP, etc. They would come complete from a Tesla gigafactory, not outsourced. And the local micro-factory would just bolt the few pieces together in a final process.
@@thelimitingfactor Currently shipping cars is volume intensive - there is a lot of empty space being moved. Relatively inexpensive mini factories where the final assembly could be done without any of the background requirements (casting or paint shop for example) could enable a great deal of flexibility. The components could be shipped from any Giga with spare capacity.
Let's do it in Canada. Canadian workers' pay is less than half Americans'.
SUPER! Thank you for going into the details of the inneficiencies of the traditional assembly processes, and all the ways that the unboxed process addresses them.
Sure thing man!
Finally, I've understood what this unboxed process is. Thanks for an excellent explanation.
Happy to hear it! Sure thing man!
this presents interesting opportunity's or trouble for collision repair. If the car is 6 major components, then the damaged sections can simply be replaced. Interesting to see how this is implemented, I see a lot of structural adhesives and SPR's being deployed
Thank you!
We all missed one important early step, executed years ago! Early Model S had an opening sunroof, with all the trim, mechanical complexity, sealing, and lots of weight. Long ago, Tesla scratched all that and began to build all models except for X with structural glass roofs, which are simply glued on! Piece of glass and glue. Can't get much simpler than that!
also Tesla eliminated the battery swap feature in model S/X too.
but , long after they closed it's battery swap locations.
@Withnail1969 sure, it's a feature many want; no argument there. However, it's a genius leap in simplicity to go with fixed glass; easier than mere sheet metal. In the context of this video, that's what's relevant.
@@Withnail1969 - #1 SAFETY, #2 it's a structural component of whole car, Tesla roof is Stronger than stamped roof, Tesla roof broke a car crusher, trying to recycle.
# 3 speed of production & reduction of multiple process.
and MORE.
@Withnail1969 what's wrong with a permanent sunroof?
@@bigdougscommentary5719 - more NVH , more weight for safety, NOT as safe or structural.
Slower production.
MORE.
Thank you so much for all you do Jordan!
You are most welcome buddy!
the Limiting factor - ❤️ loved this video, and others.
great breakdown.
@@markplott4820 me too!
You're most welcome man!
Professional and well done. Great insights.
The design of the vehicle is key to improving the build process. Many vehicle designs would be impossible using unboxed but Tesla makes EVs only with very clever structure designs so they can adjust everything as needed to get to the next level of optimization. Other automakers can't do this because they rely on parts vendors for some of the car and don't have control over it's design to some degree.
Great appetizer for 10/10.
😋
Fine video but it's rather more complicated than you think , years ago I was given a job to stop problems with trim in the assembly of the Vauxhall Frontera SUV which was being assembled from mechanical kits from Isuzu , the issue was failure to adhere the plastics due to the temperature of the vehicles coming from the "colour buffers" these were cold and were the strategic reserve for the production process , the issue caused rejects and line halts at assembly and inspection , I designed a IR heater to monitor the shell temperatures incoming onto the assembly and adjusted the heating required , it worked perfectly and allowed mixed colours of shells to lifted to the correct temperature on the fly , the shells varied by 30 degrees F depending if they were a freshly painted shell or were from colour buffer so could be days or weeks old , the colour line in the paint shop ran in batches depending on popularity and colour buffer space , there are thousands of points in the assembly that the smallest thing can cause a line delay , most of which you would think would be trivial but the consequences are enormous later down the line
Oh yeah, it's complicated as hell. I've worked in a factory. Editorial decision to keep it simple.
Wow! Knocked it out of park with this one dude! Solid video that really helped me understand the Unboxed process better! Thanks!
You're most welcome!
🙋♂️THANKS JORDAN AND YOUR SUPPORTERS,FOR TAKING A REVIEW 🧐💚💚💚
✊🏼🤠
Thank you Jordan
Another thing that stikes me about the unboxed process is there is no need to have the same number of lines for each portion of the vehicle e.g. Takes twice as long to construct the front of the car as the doors? Just have two parallel linnes for the front of the car and 1 for the doors.
An excellent point
And those stations will be less complex. So easier to modify and upgrade.
Looks like the paint shop can be a lot smaller as well..
Great thanks for exposing the un-boxed rigmarole.
I would appreciate a video about all the reasons other manufacturers haven't used an unboxed process and the ways those reasons either don't apply to Tesla or have been mitigated by Tesla.
Yes right away sir
I can give a few points, imo at least. Revamping factories is a lengthy, expensive process that requires a lot of capital and expertise. Many OEMs do not have shareholders backing them for these sorts of investments, they prefer share buybacks etc. It has also never been done to this extent so trying something new is "dangerous" to their existing business. They usually prefer the "wait and see" approach once they have reached a certain stage in their business. Also, the other companies don't have someone like Elon willing to double down when he feels the need. Plus many other factors, if the above helps I can continue, let me know. Enjoy your day. Trying to help you out if I can Jordan 😊
@@justlisten82 Thank you!
@@jessstrap2088 The problems expand when you then factor in a workforce that's also trained in processes that will now need to be changed or scrapped entirely. That means re-training or pensions and in a UAW environment neither is cheap.
@@jessstrap2088
Basically
"tradition"
Automation is simplified and easier to update with larger subassemblies.
Assemblies for different models may be produced on the same machinery. Castings may be batched for example by changing the molds only.
Great video. You fill in many of the gaps from the presentation. I only wish Tesla would use this manufacturing expertise & excellence to build cars like the XPeng MPV 9 or the Li Mega. Both have amazing aerodynamic drag for their size, more seats, middle row massage seats, a minifridge, and a rearview mirror that can switch to a screen for the rear camera. Everything people want in an EV limo or family hauler. The best thing Cybertruck has done IMO is make cars like those MPV I mentioned look sleeker and more normal. Now if only I could get the $55k XPeng in America for that price!
Thanks!
Teslas tend to be more utilitarian except for the very high-end vehicles
They've found that a lot of people don't use the bells and whistles
Most people would rather have a lower cost vehicle
Elon has commented, "Tesla can't do EVerything..."
@@FrunkensteinVonZipperneck...so Elon took on a difficult Cybertruck project while Tesla only has 2 sedan and 2 SUV (1 affordable of each) for sale. Think back to the relative ease of converting Model 3 to Y. Tesla could've built a light duty minitruck on the Y platform, assembled on the (too-often idle) Texas Y line while they tweaked & perfected Cybertruck. A small truck ready for international homologation. Instead we have an overpriced overweight Cybertruck, the ire of EU & Asian regulators, which lacks features of similar-priced Chinese MPV. Fans & shareholders can still be critics. *Should* be critics, imo.
Parallel production lines? Brilliant!
Great report.
Jordan - Tesla in 2024 a partial unbox process, and likely use Advance UBP in near future, as it works out Full UBP.
and Tesla will have many iterations of UBP in future.
In the unboxed method buffering should actually be required in sub assembly areas so if there is a slow down the extra parts made in advance can fill the gap while the problem is solved. Over time, common slowdown issues could be identified and buffer size could be adjusted to only be enough for the worst case scenario. Buffering can make the best use of vertical space like and upsize down pez dispenser.
Thanks Jordan, completely Professional as usual. The OEM’S are in real trouble and I am sure they know it. But, what to do when you have so many factories and so many staff and so much debt. Where to start? My suggestion is the Boardroom and they only get the same as their workers severance payout.
🎯
Great explanation as always 🎉
Excelent work , thanks for sharing!
How will all the components be joined? Bolts, glues, welds, tab & slot, special techniques for mating castings (aluminum, maybe magnesium) and stamped parts (steel)? Guessing continuation of Tesla's experience joining structural battery pack & castings. It might be fun to learn specifics.
The Gigacasting are directly joined to structural battery pack w/ bolts and adhesive.
Tesla has perfected wielding Gigacasting to stamped steel & extrusion.
but is constantly switching over to partial UBP. modules are becoming high quantity of cast parts , rather than stamped steel, extrusion, and small cast parts. eliminate brackets and use CRP moulded over extrusion or castings.
The major subsections will mostly be bolted together
Rover cars had a conveyor that ran at one speed only. It would have been difficult to change it. The 400kW infra red sill oven baked the sills based purely on the line speed to get the desired temperature profile, with no electronic temperature control at all.
Unboxed process allows different station process speeds throughout the factory.
I just had a thought. Is there any company which has this many, high quality and in depth, youtube channels as there are for Tesla? There is so much going on it's wild! If there are other companies with such a dedicated following I would like to hear their names. I guess Apple and maybe Meta, but it seldom feels like they are trying to solve important problems. (I love the Apple Silicon chips, but there was almost nothing out there for speculation/analysis etc. afaik)
Yes, there's no comparison 🤠
CATL and BYD are doing some cool stuff, but their presentation are always dripping with technobabble and the only really operate in a few domains at most. Tesla's pushing the limits in over a dozen different domains.
In theory the "Unboxed Process" allows for significant redundancy similar to a modern server stack where there is no single point of failure. All the sub assemblies can be duplicated and run parallel; this offers: *redundancy, *expandability & *live-upgradeability. Indeed the "Final-Assembly" can be expanded out to more than one line for the same reasons. (Sidenote: The CT door autopsy showed the modular door subcontroller with only one connector to the body - no loom. All part of the evolution towards this new process).
Tesla is almost only OEM, that uses discreet axle modules , that use high amount of casting, as opposed to wields, fastener, brackets, and different materials, like stamped steel, extrusion, and small cast parts.
the entire axle can drop out to be worked on, just a few bolts.
Yes..... The best step is NO STEP!
A big problem was what if they need something repaired. Not having many repair stations means praying nothing happens or make it easy to fix if.
The first time I saw a Tesla factory, it was a shock to me. For electric vehicles, Tesla is undoubtedly the best choice
In Munro's rear down of the cyber truck, we saw quite a bit of the unboxed process, particularly the door modules
11:27) Jordan, "casting process" is necessarily asynchroous: faster casting is required to store part inventory to be used during required downtime - for switching out dies and maintaining GigaCast machines ...
Good point!
It also gets me to thinking about how all these parallel lines will be asynchronous in terms of speed
Great video!
Excellence
In a spacex star factory tour Musk used the term "liner adjacent flow"
Thanks!
Thanks for the support!
Great video, Jordan. You did not mention Tesla One, the overall manufacturing management system. Which manage also spot changes and manage necessary relevant changes on other systems. This will be much easier in the unboxed process. It can also ensure that buffering mainly occurs in the inventory section and that the line itself hols the optimum amount of sections needed to complete the part. It may even be that the unboxing proces spreads out to the other parts of assembly line. Like for instance building the structural battery back with additional elements like chairs, suspension etc. Tesla can provide information to see if it is more efficient to have an element move to another part of the subassembly. Exciting times.
Thanks man! Yeah, there's a WHOLE lot I didn't cover here. It could have been a 2 hour video easy.
Overall good additive points. Thanks for the comment.
@@thelimitingfactor Yeah, so True. If you decide to do a deepdive on Tesla One, maybe do a co production with Joe Justice. He seems to be on top of this.
Thanks for the deep dive into unboxed process. Answered a lot of my questions.
The pre-assembled sections will be _bolted_ together, instead of being welded. Pretty wild!
About 40 years ago Volvo developed and implemented a similar process as the proposed Tesla unboxing process.
Volvo did not continue with it and returned to the traditional production line approach.
Tesla has the advantage of gleaning lessons learnt from the Volvo attempt as well as having the advantage of more advanced technology being available today.
🤞
These ideas are rarely new. I wonder how franken-car stitched from 10 parts will look and behave in long run. Of course, every car is built from parts, but each part of process allows for alignment to happen at end of that point. I guess consumers will find out in the end.
I suspect the advances in automation and manufacturing precision in the last 40 years will make a big difference towards making the process more viable. something like the unboxed process requires very tight tolerances to go together properly.
The gigacastings with integrated brackets instead of lots of separate parts also dramatically reduces compounding errors, so that's probably a big part of what makes the unboxed process even possible even with modern precision and automation.
Amazing explanation as usual. These tech innovations synergize in interesting ways, now I can only imagine that if Optimus were added to reduce humans in the unboxed production lines, optimizations, speed increase and cost reduction could grow even further. Seeing how 10/10 logo says "we robot" I can only hope all these improvements will make it possible to build not only the robotaxis but also the affordable $25k compact car as well. It's not only that people in the world need those to exist for the EV revolution to keep happening, it's also that detractors arguments that ICE cars are still cheaper will be squashed and won't stand anymore, so it must happen, it must be done, the world needs it to happen and everyone who trusts Elon is hoping he can make it happen, as he has shown he is the man that makes the impossible, possible.
💯
Tesla AI factory 🏭 software is underappreciated.
every Tesla made , has a digital twin in software, and in 2024 its final assembly is done w/ computer installed, and running in Service mode. this is to reduce mistakes & alignment.
Yes! There's so much more there to cover!
I like the production design and the new technologies for the Cybertruck. And I like the concepts to be employed in the unboxed process. But my hope and question is whether and when these advances will be applied to the Model 3 and Model Y (not just for CT or suggested future vehicles). Maybe that would be a long while; require new factories (or perhaps there is room in some existing factories for new assembly lines); or maybe actually never for these models. The "trouble" (odd to call it that) with EVs is the rate of technology advancement. Like computers, both the factories and the products are almost obsolete when you buy one. About the only thing that can keep track of that (and assist when service is needed) is the software that defines and runs both.
Yeah it's going to be a long time before they do this with the model 3 and Y
There is one big change we haven't discussed and that is how to fix the different vehicle modules once they are sub assembled. Welding is probably not an option. Using screws is challenging as they have to be many or there will be stress concentrations. I would guess glue plus screws would be the best. On the Cybertruck with huge castings and very strong panels, screws are sufficient.
They covered this in a patent about a month ago
I linked it somewhere in one of the comments below
Bolts and adhesives from memory
Nice!
The Ford Model T was produced with a process fairly close to the unboxed, however that required a separate frame which is a big drawback for the final product.
I always imagined that by having sub-assemblies it enables multiple Vehicle Assembly lines in parallel. Current lines are 30-40s for a car to be completed off the line, but instead have multiple assembly stations (rather than a line) to assemble each car... wont need snaking sequential line running through the factory but rather just nodes or almost pitstop like areas where the assembly is done and drive off?
So Tesla is risking the paint colors not matching as well and either skipping spot welds between modules. In return they get to decouple the design of the 6 major sub-assembly lines.
🎯
The traditional reason for painting the fully assembled body instead of individual parts is to produce a more even and uniform paintjob, especially with fancy metallic-effect-paints that produce shiny reflections and patterns that are supposed to appear uniform across the entire car.
Depending on available paint options that may be something to think about.
Yes, I had heard that painting variations could be the Achilles heel of the UBP.
Definitely! Who knows, maybe they'll use stainless steel
At most that would just result in fewer paint color options. The trend recently has been a move away from metallics anyway.
@@thelimitingfactorI think powder coating would be the fastest and most consistent as it is applied dry and the smaller parts could then go through a continuous oven.
The oven could be as small as 4’x5’ opening…. Then again no paint would be best.
Making the seats in house also saves an absolute ton of logistical nightmares, as well as reduced shipping costs.
Tesla unboxed FTW!
In 2024 oems are Still having to rework stamped parts with 2x4 wood, and 🔨......lol.
Gigacasting requires zero rework or post treatment.
Cast parts fit perfectly, everytime.
Didn't Morgan use a wood frame? Cute cars. Scary as Frunkenstein, but like Frunkenstein... cute!💙
What I don't understand about the unboxed process is that it is said to take less factory footprint area.
It seems to be that the unboxed process is similar to paralell processing in a CPU which utilises more die area, though admittedly here more throughput is achieved via multiple channels whereas with the unboxed process the same amount of assembly is carried out in less time.
It seems to me that thought the lengthe of the main assembly line will indeed decreace it's 'width' will increace.
Does the unboxed process really take less factory footprint area, the same, or more?
I'd have to re-look at the presentation, but at minimum, it's creating more output per square foot of factory area by accelerating throughout.
@@thelimitingfactor Thanks for the reply.
I still have my doubts that it even creates more output per unit of factory area however.
I can see that it reduces the time from start to completion due to improving ergonomics and parrellisation but if you have to provide workstation areas beside the production line this surely must require more space. The length, (and area) of the main line will indeed be reduced however.
Be careful of conflating the concepts of rate of throughput with rate per unit of area.
@@philflip1963 "IF you have to provide workstation areas.." What IF you don't have to...?
@@FrunkensteinVonZipperneck Surely you must have to provide workstation areas, either on or off the line, otherwise the work cannot be done.
So... my imagination got the better of me and I pictured making hamburgers.
Current car assembly lines are like this: The burger bun is baked as one piece, then buttered. It is then sliced in half and separated, and then carried on a conveyor through the length of the building. Meanwhile, the hamburger patties are carried along, just an inch above the bottom bun, and tiny workers climb in the gap and spray condiments on the bottom bun. The top bun is only 2 inches above the meat, and workers use robots to bring in the lettuce, shove in the onion, and try to stuff in the pickles and tomatoes without them sliding off the outside. Then some crazy South African comes along and says, why don't we cook top and bottom bread rather than a bun we have to bake, cool, slice, butter, and carry the length of the factory? We can have 2 ovens, one on each side. Then we bring in the bottom bread with condiments already applied, meat with the veggies already sitting on top, then stack the top bread. Stacked all up, made fresh and more quickly! Think outside the bun, y'all 😂
So.... My imagination got the better of me and I thought....
.
.
.
.
Mmmmmmmm.... Burger!
@@rogerstarkey5390 it's what's for dinner!
How does one butter your Model B before it's sliced in half? You obviously have no autoburger experience.
So elves are gonna take over as Big Mac fry kooks?
What's your thoughts on Solid state batteries? Especially since MG claim they will come out next year. And comparing to Tesla's battery strategy. Could make an interesting video?
We can hope, but so far they're a lot like Fusion, always a year or two away. I think Toyota has been planning to produce solid state batteries next year for going on ten years now.
At least for the next 10 years, solid state batteries are going to be considerably more expensive than batteries using a liquid electrolyte
This makes me want to play factorio.
🤣✊🏼
Great breakdown, Jordan. Hopefully Tesla gives us more insight as the line is finalized.
Maybe we’ll learn more at the robotaxi event. They’ll unveil the new design, and possibly go into the manufacturing process for it.
Tesla will NOT give away ALL it's secrets to reporters, investors, or consumer.
Tesla has technological LEAD over every other OEM on EARTH.
@daveinpublic new design, check. Rough start date, but no real details on manufacturing. Up to the Tesla inteligencia to fill in the gaps again
Tesla is transforming car manufacturing. As sognificant as Henry Ford on the Model T.
By making the paint shop more efficient, they also lose a major hidden issue. Paint.
Automotive paint has to be stored within a tenth of a degree. It’s how a lot of long term warranties on body’s are achieved. Less paint less storage, less energy.
Better paint efficiency also means less pollution. VOC’s are a problem for paint shops to control, and those controls are maintenance heavy.
Often they use wet booths, however Tesla have already started to attack this point with a dry system. So less water usage.
Paint Shops are Frunkin' HOT! Generating all that heat requires huge cooling costs. A vicious, expensive cycle...
EVs are basically a fisher price power wheel for grown ups, I expect them to be exceptionally efficient to assemble.
🤣💯
TOPIC SUGGESTION: If a battery is cooled from the top and the bottom, can you create a battery 1000x80. Like 1 meter diameter 80 mm tall big pizza sauce coin cell battery. It can be used for grid storage only.
no
Volumetrically, it makes more sense to use cubes
@@thelimitingfactor I suggested this for stationary storage where volumetric and gravimetric efficiency don't matter as much. What matters is the ratio of active to inactive/structural material.if you have a stainless steel barrel and inside you have a stack of jelly rolls separated by cooling plates(disks) that also act as current collectors. This approach will improve manufacturing speed and reduce cost
Nice animations. Had to speed it up to 1.25.
👍👍
How does bolting the sub-assemblies impact the crash performance?
Minimal impact. Sorry I couldn’t resist.
Actually won’t change anything as the whole assembly is the tested piece.
It won't, you just have to design to it
Are you sure? Modern vehicles use adhesives to augment stress concentration areas ( welds, bolts). E-coat oven cured adhesives are higher performing than ambient temperature cured adhesives. Unboxed requires ambient temperature curing.
They should do an unboxed process
Thanks for this, nice summary. Not mentioned were steer by wire and brake by wire, implemented in the Cybertruck. This is enabled by the Canbus loop, that you did mention. Removing steering wheel shaft and hydraulic lines for brakes helps the unboxed process by removing difficult linkages between sections of the building. Enabling a plug and play assembly.
Thanks! That was covered in the last video of the series.
What are the cons? Could it be that less parts then it's more expensive to replace each part? If the car gets hit behind the back, it doesn't seem to be an easy fix
No cons
It just takes a much higher level of technological skill
When they assemble the finished parts, and they don't align within tolerance, maybe it would be harder to get them into alignment without damaging paint/finishing. I hope they'll be designing the fastenings to be able to handle this instead of the cybertruck "make it all be micron level accurate", and when it isn't you're screwed.
Other "cons" is that this is a new/untested process. We don't know how well this will end up working in practice. That goes for any innovation though.
Crash worthyness and lifetime may be negatively impacted by this, since the overall structure of the car is somewhat different because of the fastenings.
These aren't unsolvable, but new is different.
👍
are we going to talk roundabouts with pedestrians crossing countdown process…
#boxjunction
17:00 the doors have their own controller
The point was to show the general architecture
keep in mind, the WORLD will go Gigacasting, structural battery pack & unbox process.
Cherry auto in China, now using 16,000 ton Gigapress to make complete Underbody.
and TK press makes 9kT, 12kT, and 16kT presses.
the 16k ton gigapress, is same size as apartment building on its side.
But a completely cast underbody doesn't allow for the easy access of an unboxed approach.
tesla uses 9t press becouse the fluid dynamics goth probably optimized to the max and beyond. also less pressure means longer lasting tools i hope
The question is can this produce a quality product... its one thing putting together a truck or a bottom of the barrel cheap vehicle, and another thing putting together a high end product where long term comfort and ease of use, and very low fault tolerance, are the norm
It already will be improving quality. Giga casting creating a solid frame for the panels to fit to. The battery and seat sub assembly cutting down on through door mounting damage risks.
.... consider adding brake-by-wire... ?
Covered in the last video
In case one of the lines takes twice the time you could just go with 2 lines?
Yes! I should have covered this in the video
@@thelimitingfactor i am wondering if any process in the traditional manufacturing is a bottleneck- if so I assume this kind of „optimization“ would help to increase production speed further
I see Optimus robots working on these sub assemblies lines to further decrease labor cost!! Or am I crazy?
Yes!
Can you say Robot? We are killing ourselves with automation. Tesla may unionize to protect employment. I don’t blame them.
I'm not sure what you mean. Virtually eerything around us is thanks to a certain amount of automation
In most large organisations, the dooers and innovators get weighed down or stopped completely by the parasitic departments and functions that spring up around them. Many young people walk unto these companies with fresh ideas and are crushed by the corporate machine that rewards mediocrity and greasy pole climbers more than with long term players trying to help the company.
🎯
If you list incompetent, spanners-in-the works, please give a Taylorism shout-out to Human Resources!
As usual 👌😊❤️👍
If the body sections are put together later the how/ what method to weld together through the paint? We need to know!?
They're going to be bolted together
If you have precise location points, for instance a pin on one part fitting a hole in another, you will get a far more precise and controlled structure with minimal cumulative variation.
Then, as Jordan says, you can bolt them together.
Add that to the precision of the megacastings and the overall tolerance of the complete vehicle dimensions should be impressive.
.
Jordan.... Another advantage of bolting together?
Major repairs?
Let's say there's front end damage, severe enough to require a Casting change (I know.... Unlikely not to be written off)
.
The whole assembly process is *designed* around simple and rapid connection between sections of the vehicle?
As such logic would suggest easy mechanical and electrical *disassembly* of the front casting "unit" from the main body, possibly allowing for that major repair by bolting on a new casting preassembled with old and new subcomponents as required rather than "bending" a conventional body to shape.
.
This may in fact lead to FEWER vehicles written off?
They said castings would fail too.
Does this unboxed process make outsourcing to outside contract manufacturers easier?
Outsourcing the building of their cars?
No
But that's not something Tesla would do anyways because it's core to the business
Outsourcing the one thing they don't want to do. Have you not heard about vertical integration and how outsourcing has delayed and causes problems in the legacy lines? It's worth your time to do some reading if you're interested.
no, outsourcing is NOT the future.
IN - house sourcing, is the future.
I think this opens up more options for the subassembly lines. One subassembly backing up more than the others? Could put in two subassembly lines for that.
This would be capital inefficient if I'm catching your drift correctly
@@thelimitingfactor
More likely that if you had let's say the rear cast and battery/ interior section running "fast" you would simple stack the extra units and pause those sections.
you don't need paint line for Cybertruck.
Like, when they "unbox," how ya gonna recycle all that cardboard?
😁
why not painting insides? tesla can rust, and any aluminum alloy can be real bad with it. Some people used tesla cars in high humidity countries ended with rusted through areas (as they said - turned in powder). Tesla car shouldn't last for 5-10 years, that'll make it's most expensive part - battery pack - useless (that can last for 30-40 years)
Hmmm, I haven't heard anything about that about the rust. I'm curious which years that was for.
As for the battery. No, it won't last 30-40 years. More like 15-20 depending on the chemistry.
@@thelimitingfactor rust goes from wet dirt that that have some salt in it. PArts that involved most in it is places behind wheels (mostly, front). Thin anti-corrosive layer goes down from 5-10k miles(maybe faster), then dirt placed there starts the rust and if that is started, there is almost no way to stop it since it goes deep into metal.
Batteries are just guaranteed to use in 15-20 years, they can last a lot longer (and there's no real risk of car self burning if chemistry is LFP) so that cars and used battery packs will be there, it's only matter of time.
One possibility you didn't explore is having redundant sub-assembly lines and some permanent buffering of sub-assemblies. This would allow production to move to the backup while a line was down and keep up the vehicle production while a sub-assembly line was being addressed. It would only make sense if there were multiple lines for each sub-assembly and if stopages were common enough. Less efficient than the designated sub-assembly line in operation, perhaps, but still maybe better than a main line shutdown.
This seems very capital inefficient. Unnecessary redundancy.
Creating parallel production lines gets rid of the bottlenecks and makes main assembly simple
It atomizes and modularizes everything which reduces issues
@@thelimitingfactor Having a main production line down has an hourly (actually, often by the minute) cost. Wages are often still payed, at least in the short term. Overheads continue unabated and cost of capital is undefrayed. While having "just in case" storage and production is rarely cost efficient, a small amount of redundancy/storage can be the most efficient by avoiding these stoppage costs. Indeed, Toyota found this out quite a while ago and aims for "most efficient working capital", not "lowest possible working capital". The US majors didn't pick up on this, which is why, initially, Toyota fared better during the pandemic.
As an example, I heard of one case where a vital part in a car line died and the company flew the replacement in for around $100,000, many times the value of the part. However, the failure was so rare that it didn't pay to have a replacement everywhere the part was used and the cost of a stopped line was so high that the cost of the flight was lower than the cost of the line being down while the part was driven there at far lower expense.
The optimal approach is, I suspect, a linear programming problem dependent on the frequency of line failures, the average length of time to fix, the number of identical lines, the capital cost of a redundant line and so on...
I think the idea has been abandoned
No, they've confirmed it's still a go
unbox process is the future, and AI factory 🏭 software.
almost no one else uses , warehouse on wheels in auto manufacturing.
and many years ago , Tesla bought Rivera tool & die. and Canadian assets. Tesla can make ANY tool & dies it needs for production.
Take a shot everytime has says "unboxed process" 😂
You vocal pitch variation sounds unnatural
Not sure what you're expecting me to say here. That's my voice.
Might be difficult to get the paint to match on individual parts ..Having owned 3 Tesla's , Tesla has not been good with paint.
I agree that will be a challenge!
Maybe they run the parts for each car through the paint shop at the same time but not assembled..
Perhaps they've gone from being below industry standard with paint to figuring out how to be above standard and more consistent such that parts don't even have to be painted together. One can hope.
@@thelimitingfactorThe paint match is why the doors are mounted first and removed later in a 'boxed' assembly process. Solve that problem and a 'boxed' process can do away with that inefficiency as well. Nothing to do with boxed or unboxed.
@@jamesengland7461
They've not been "below industry standard" for a long time (maybe Fremont isn't as good as Belin/ Shanghai?)
Rover cars had a conveyor that ran at one speed only. It would have been difficult to change it. The 400kW infra red sill oven baked the sills based purely on the line speed to get the desired temperature profile, with no electronic temperature control at all.
Unboxed process allows different station process speeds throughout the factory.
Excellent point!