Jordan is an absolute treasure. He simplifies complex technologies for us to grasp. Tesla is so far ahead it is hard to truly understand the revolutionary developments without Jordan’s work. Thank you.
IMO, one point of steer by wire that is largely overlooked is the fact that still large parts of the worldwide market is righthand driving. Common parts and processes can be greatly increased between left hand and right hand models - ifyou have electronic steering, breaking and accelleration.
@@thelimitingfactor The steering/ pedal box can be essentially a "Game controller" ("Tesla, controlled by Logitech"? 😉) . Great for a common Compact/ Taxi design. Build the Compact first, *Lease* as a driven cab. Wait for regulation. . When approved, THAT NIGHT, delete the modular control box from the line, change firmware and you're building CyberCABs. . Take the stock of steering modules to the next jurisdiction (Europe?) and repeat.
One other advantage of steer-by-wire and brake-by-wire is that you can have a vehicle that has a wheel and pedals, but when in robotaxi mode, it can disable the controls, rendering it immune to interference from passengers (and it still has the controls for manual operation, either when just in manual mode, or when the software is unable to handle it)
My company MAEC (modular affordable efficient car) invented the modular/unboxed process in 2007. I just didn’t have the financial backing to make it happen. Electrification is the real key that allows simple interfaces.
Super exciting that we get to follow this developmen tin realtime. The final product is truly the future. Thanks for the video! Your dedication to be precise and cover benefits as well as caveats is appreciated. Easily recommendable to other curious parties =)
Steer by wire makes it easier to make cars for both left hand and right hand drive markets as the steering mechanism is the same for both types. The only real physical difference between the two configurations is swapping the locations of the steering wheel, brake and accelerator with the glovebox in cars like the model 3 and Y.
Just hearing Musk saying parts of the unboxed process will be used on current production lines is exciting. Seeing this tech in use is when theory moves into being real. I agree that if the system is fully developed, it will be a huge change for industry. Keep up the great work Jordan!
Something I’d like answered. 12v systems, even with larger wires, have disastrous corrosive reactions in high-salt environments-wires and connectors literally falling apart. Won’t a 48v system with thinner wires and larger voltage differential have an even more likely corrosive failure potential?
It’s like building rifles; upper receiver, lower receiver, … it enables a whole secondary market to tweaking and providing modules. The legacy car manufacturers were the ones integrating many sub modules from suppliers; Tesla adds an extra layer of modularity, enabling more to be parallelized and iterated separately. It will increase development iteration cycles inside modules 🤯
I like how although there are many more redundant parts that there is the potential for the car to specifically notify you of exactly what is wrong when something does fail so the repair can be very exact.
Unboxed process sounds like the way to make everything easier while reducing costs and complexity. They could even use it to increase production for all the other vehicles (S3XY) by adding an unboxed production line for those. Also, knowing that Optimus will start working at the factory soon, I can imagine pretty much all the work could eventually be done by robots. Elon's dream of a gigafactory that looks like an alien dreadnaught will be fulfilled. But what I'm curious about is the thixomolded magnesium gigacasting you talked about in former videos. I am hoping it would be introduced alongside the unboxed process as it would imply another important improvement in manufacturing by reducing the weight of the cars!
@@seancollins9745 I thought the alloy that was being developed solved all those problems. And anyways, even in the worst case scenario, wouldn't it be possible to make the Gigacasting of the battery pack of Magnesium to reduce the weight but keep making the front and rear gigacastings of aluminum to keep those crumple zones the way they are, but that would be only if making them of magnesium didn't work, which I doubt.
@@thelimitingfactorI too normally rewatch your videos 3-4 times over a couple of days as they are jam packed with incredible insights and very detailed. I wonder if Tesla will go exploring magnesium Giga casts over time? That’s a different kettle of fish!
@@thelimitingfactor It's ironic. The quality of your narration is particularly consistent. Computer generated voice is not as good as high quality human narration as yours is, but computer generated voice *is* consistent, bordering on repetitive/monotonic. Nowadays people take that consistency as a cue that a narration might be computer generated. You can't win.
I believe it will help repair by allowing to take totaled car and reuse sections for a new Frankenstein car. I’m also thinking maybe they find a way to share fronts between models; model X, Y, Van, and Non-Cyber-Truck. 3, robocab, and maybe a cabriolet. So there could be some system of LEGO like fronts; heavy, high, undercab, narrow, big frunk etc. then different side profiles, different rears, and different top+bottoms.
Brakes don't necessarily need to be electric, at least in the front. Just like the steering wheel, it could be fully contained in the front module. The rear, on the other hand, wouldn't make sense, as mentioned. But rear brakes are smaller than front brakes, so another option may be to eliminate entirely, and just brake regeneratively (especially in the Robotaxi). "The best part is no part" philosophy makes this worth consideration.
VW uses rear drum brakes in addition if the goal is to maximize motor braking a resistor mounted to a heatsink could be used as a place to dump the remaining energy. It would then be vented out of the car by a cooling fan.
This is an amazing analysis. All of these upgrades could have been developed by the old automakers but they just refuse to innovate... If more companies were like Tesla humanity would be already in the age of abundance
2 advantages. Castings. The ability to machine dimensionally accurate mounting faces for extremely precise location in X,Y, and Z axes. No accumulated tolerances. . Etherloop. Main(?) safety advantage? . Its "a loop". A single, but complete break in that loop still leaves every individual module a connection to "home". If you have a controller at "12 o'clock", and modules at 4+8 o'clock A break anywhere in the ring still leaves at least one clockwise or anticlockwise route to "12". A single test wire in the loom connected to different pins could detect a problem and advise service required, but not disable the systems.
The 10kt gigapress is big enough for door ring assemblies, let that sink in😂 The hydraulic brake lines could just use unions, many linesets on cars already do come and are assembled in multiple sections, usually on unibody cars there are lines on the subframes and lines on the body side, i don't think electric brakes are a requirement for unboxed construction
I have a comment to the Tesla Etherloop wiring video. I wanted to submit this comment to that video of course, but I was late, and by that time you were no longer reading the comments to that video, by the looks of it. (No criticism there, you have to stop at some point.) In cybersecurity there is the concept of "air gapped" A mission critical system should not have network connection to other components in the same facility. The obvious example is a nuclear plant. The offices that handle logistics must have internet connection, but for security reasons the computers running the nuclear reactor must under no circumstance have network connection to the rest of the facility. If the office computers are hacked, that is bad enough. But as a bare minimum the nuclear reactor controllers must operarate on an island, no network connection to the rest of the facility. The reason for keeping the nuclear reactor controllers on an island: if a target is very attractive for an adversary then the adversary will allocate resources towards hacking that target in proportion to the possible wins. The thing is: try as you might, it is impossible to write unhackable software. There are diminishing returns. Even if you set up tons of quality controls, that make the software 10 times or 100 times more expensive in manhours to develop, you can still only make the amount of errors smaller, not zero. It is extremely irresponsible to go with a design of cars where life critical systems and entertainment functionality are sharing a network. Promises that the critical system communciations will be encrypted don't cut it. Try as you might: it is impossible to write unhackable software. Mitigating efforts: Like many other companies Tesla offers large financial rewards to white hat security experts who develop and demonstrate security breaches. These experts do succeed in finding ways to get high privilege access to Tesla systems. It is to be assumed that enemy nations are in the process of stockpiling ways of access that they do not disclose to anybody else. (Such stockpiles are always in flux, vulnerabilities that are discovered and reported are patched, so new ways of access must be found and added to the stockpile.) When white hat security experts find a way it invariable starts with gaining high privilege access in the entertainment functionality. That then forms the platform to subsequently find a way to access and gain high privilege access in the life critical systems. A fleet of cars with shared critical systems and entertainment system is for an enemy nation a sitting duck. About over-the-air updates Tesla has a policy of over-the-air updates. That policy puts a huge responsibility on the security of development and distribution. For comparison: in 2017 the utility software ccleaner suffered a hack of their distribution servers that went unnoticed for weeks. For weeks a particular ransomware was distributed with every update of the ccleaner software. If the Tesla distribution servers are ever hacked then in minutes all of the Tesla cars can be infected with malware. When the distribution of updates is not over-the-air then the update cycle is a much slower process. The only practical option is that updating the software of the car is performed when the car is brought to a shop for periodic maintenance. Slower distribution is beneficial for security: better probability that if somewhere in the custody chain the software is compromised the tampering is discovered when the number of cars infected is still comparitively small. Jordan, I hope I can persuade you to allocate time to research these matters, and to communicate to your audience what they need to know to make an informed decision. The reputation of your channel is that you are actually researching your subjects, rather than repeating the content of press releases.
as long as the system is robustly tested I am not too concerned with electric brakes. There is no redundancy in hydraulics if you spring a leak all your brakes stop working. With the ther loop you'd need two lines to loose communication before you have a failure already making it safer. Also the duty cycle of brakes is much lower on electric cars which should increase reliability over a given time because of the lower use
"There is no redundancy in hydraulics if you spring a leak all your brakes stop working. With the ther loop you'd need two lines to loose communication before you have a failure already making it safer. " There's redundancy with hydraulic brakes. You would have to puncture both circuits in order for that to happen.
Excellent video. Particularly the discussion of electric braking. For almost all "normal" driving (e.g., not "off-road"), computer control of braking (for regen or for anti-lock or perhaps for skids or wheel-slipping) offers much greater safety and fewer parts, etc.
Hi Jordan. I haven't heard it mentioned yet, but in-case brake-by-wire couldn't get approved, it's still possible with hydraulic brakes by using quick disconnects. There are refrigerant QDs to disable part of an HVAC system, even though a very small negligible amount of air is introduced upon connection. There is no reason that a hydraulic QD couldn't work as well. Each sub-assembly could be either primed or under vacuum to accept the fluid when everything is connected.
This may be a pet peeve, but I don't like the term "reduce complexity" - a lot of these decisions move the complexity earlier in the process or inside a part. Internalizing the complexity DOES aid in final assembly but it does not get rid of it.
We've heard the analogies about aircraft having fly by wire systems for decades, so we needn't worry about cars having the same system. Another aspect is the sheer size of these aircraft, whether they be a fighter jet or airliner. The forces needed to move the controls are far beyond what a human can generate, and as much as 200 feet from him. You can safely drive a car if the brake power booster or steering power assist fail, but you're not going to be able to fly a 747 by hand. Brake by wire will work just fine, as it already does with electric controls on aircraft. Hydraulic lines merely transfer the boost force from the power assist over the distance to the brakes, where with electric systems the boost assist, commanded by wiring, is attached to the brake disc or steering rack. You can easily have pairs of motors, and even if they're used in tandem, the loss of one merely means the loss of half your boost, not all of it.
F1 cars use brake by wire systems, you know the fastest accelerating, hardest cornering road vehicle's made by mankind. The now 200+mph capable formula e gen 3 cars this season eliminated the entire rear friction braking systems relying instead on brake by wire all around and the rear motor itself for all of it's braking done through regen only! Also throttle bodies and steering systems have been by wire for decades now to varying degrees. 😀
I’ve been waiting for people to dive into the unboxed process. Thanks for covering it! Joe Justice did a great presentation on this in Paris last year. Besides his videos, there’s not much information about this crucial part. Especially when you consider the potential of Optimus assembling the modules. If we can get the module components down to weights Optimus can handle (currently 20kg), it’ll be much more cost-effective to retrain an Optimus compared to using costly, static robotic arms. This is likely what Elon means by Gigafactories being “the machine that builds the machines.” Love your content! Check out Joe’s presentation if you haven’t seen it: ruclips.net/video/ah2O76h-c_Q/видео.html
I hadn't considered the advantages for unboxed with electric brakes. Also, question. Previously, Tesla had a patent for "rigid" wiring as an aid to assembly. I can't see any reason the ribbon wiring can't be rigid, do you think they will introduce that change as well in the unboxed process?
I think it comes down to use case. In some places rigid makes sense and others flexible might make sense. For main power lines in the high voltage system, it's already rigid
I'm no electrical engineer, but couldn't the engines also be used as active brakes? So not just regenerative braking, but reversing the current when needed? Also, when adding a capacitor to store some of the energy might reduce the load on the battery pack, while allowing this regenerative or even active braking when the connection to the main power is lost. And could such a setup not completely eliminate the need for disk brakes?
Someone suggested using the motor case as a resistor. I'm sure there's some sort of creative solution as you're suggesting. But, that's beyond me! Electronics cook my brain
I'm no electrical engineer either, but my best guess is that there are possibilities. When the motor is used for braking then kinetic energy of the car is converted to electric energy. If it was just up to the motor: I think that the motor can deliver the same braking performance as acceleration performance. The electric energy that is generated during hard motor-braking has to go somewhere. The limiting factor is ability of the battery pack to absorb that energy safely. It could be, I don't know, that an attempt to charge the battery pack faster than it can absorb will result in runaway overheating, and then the whole car will burn. But what if you had an additional pathway for dumping energy? When the model 3 was introduced it was described that the temperature management system has the following capability. In cold temperature the motor can be controlled to run differently, in a way that generates a lot of heat in comparison to the amount ot propulsion that is generated. Usually you want to minimize the amount of heat that is generated, both during acceleration and during regenerative braking, since heat is wasted energy. But in the case of an emergency stop it would actually be beneficial if the motor generates a lot of heat and comparitively little electric energy. Then quite possibly the rate of production of electric energy produced is within the ability of the battery pack to accept incoming power, with the rest of the energy going to heating up the motor. After the emergency stop the temperature management system will have to start work to bring the temperature of the motor back down again. Electric cars can do that: the motor does have coolant loops. The motor has high efficiency, I think something like 97 % of the electric energy is converted to propulsion, but at highway speed that 3 % does add up, which is why the motor does have coolant loops.
Re: Your comments on braking. The motors can create reverse torque in an emergency. I think they already do! I had my car initiate emergency braking when a deer came into the road. I was already on the brakes fully (or nearly fully) when the emergency system kicked in. I'm ~99% sure the brakes weren't getting the job done and the car activated the main drive motor in reverse.
Nice explanation. Rather than routing the ethernet and power cable after the assemblies are joined together, couldn't they have connectors at the margins of each assembly and then simply clip them together when they are joined? Another question: I presume, as you mentioned, the assemblies will be bolted together rather than welded because the welding would damage the post interior finishes and paintwork. How will bolting compare with welding for strength?
Presumably after most accidents, the car may be partially "unboxed", the damaged unit replaced, and then "re-boxed"? Or maybe a sub-assembly could simply be removed and replaced directly?
I wonder to what extent the lifecycle emissions associated with each Tesla vehicle will be lowered by the unboxed manufacturing process, in-house manufacturing of 4680 cells, vertical integration of lithium refining, and the replacement of all ICE vehicles used by Tesla and their suppliers with Tesla manufactured vehicles. Currently, it takes three years for the additional carbon required by a Tesla in manufacturing (compared to equivalent ICE vehicle) to be offset during use of vehicle. When does that become zero?
I'm sure Tesla said it was less than 3 years (improving!) . Vertical integration, if you can get it into 1 factory, reduces component transport cost/ emissions. Then, use the SAME components across ranges to compound that. . EVENTUALLY (the BIG one) . STOP building new cabs. Return them to what USED to be a "Terafactory" for *Refurbishment* Remove parts, units, assemblies and rebuild those inside the factory, fit previously recycled units to the body/ battery (50 years?) and send a "remanufactured" cab out the door. . Reprocess the waste as raw material on site.
Who said 3 years again? I haven't seen anything peer reviewed claiming anything like that long recently done. 🤔 The latest I've seen show evs as paying the additional emissions back in 8,000-18,000 miles depending on the class of vehicle discussed, pack size, location of the owner, grid mix emissions and who manufactured it. The average tesla sold is under 11,000 miles in the U.S. based on our average grid emissions for break even. Less if you have your own solar panels or a high mix of nuclear&renewables available. Plus the grid is greening itself month over month&yoy at break neck speeds! 😎
As far as I remember it was three months (or six). It was not years for sure as the energy consumption for the production is already only marginally higher than that of a comparable ICE car.
There are at least 4 major ways to get to zero additional carbon 1) Increase the operational life of the vehicle (use long lasting materials like stainless) 2) Increase the operational duty cycle of the vehicle from 5% to 50% (roboTaxi) 3) Make the cars rapidly refurbishible. (replaceable batteries, cushions, etc) 4) Reduce the likelihood of loss of vehicle in a crash (FSD)
I'm not sure the drive by wire feature has any cost benefits for smaller vehicles and wonder if that will be left out of most future vehicles. The 48V Ethernet archictecture seems like an extermely important evoluation for cost, reliability, and manufacturing ease. I realize you can't have drive by wire without the 48V architecture, but you pointed out that the cheaper steering rack can still be used with unboxing process. Mabye the new Roadster needs drive by wire for better control at high speeds, but will Model Y and 3 sized vehicles need it?
@@gordontyler9317You are forgetting a major point. Manufacturing Scale. Just like chip makers turn off functionality to get a new skew and therefore have just one part to manufacture... Tesla can quite easily have every single model use the same Steer-by-Wire unit. Only the rack would change for different geometries and sizes. You can easily save enough to put Steer-by-Wire on every car.
I think the electric brakes could rely almost entirely on using the electric motors. Maximize resistance during regenerative braking to the point of stopping the motor.
Always puzzled me why generative (as opposed to regenerative) brakes have to be limited by the capacity of the battery to absorb the energy. Just put an electric heater in the thermal control loop and dump any electricity the battery cannot absorb into that. This is what friction brakes do anyway, turn kinetic energy into heat, but this way there is no brake wear and sometimes the system will be able to put the heat to good use for heating the cabin or preconditioning the battery.
Hydraulic quick connect unions would also be compatible with the unboxed manufacturing system and cost way less than motors on the wheels as hydraulic calipers are small, lightweight and reduce unsprung weight as oppose to having motors onto the wheels. The unsprung weight has major effects on vehicle performance. There is no valid reason to eliminate hydraulic braking systems when making the the unboxed car. The plumbing would simply be broken up into sections and joined with quick connections, then the completed hydraulic system would be bled just like any other vehicle! Don’t overthink it Jordan! There are ramifications you aren’t considering, when it comes to braking systems.
@@thelimitingfactorHi Jordan. In your recent podcast for Farzad, you mentioned that Tesla said at their investor conference that the affordable Tesla wouldn't use the whole Unboxed process, but here Cybertruck uses 6 out of 7 pillars of it, and you said that Model 2 could use brake by wire because it's lighter. So what's stopping them from going all in on Unboxed Process could you explain?
They're not allowed to give feedback, but they do watch them The leader of their 4680 program did make a comment on X saying I was the most accurate source of information
Brake by wire? MAY be an offshore of regen, in which case, positive impact? Steer by wire? Tesla already uses dual steering servo motors capable of FSD, so NO input from the driver.... So no (extra) impact . And the power requirement may be lower
@@rogerstarkey5390 For Cybertruck steering, every little bit of driver input, every little micro movement has to be driven by the steering motors, they take a lot of power as steer by wire was only possible by 48V. Maybe that’s partly why range of the Cybertruck was woefully disappointing. Tesla vehicles already use regen braking but they still require mechanical braking from the hydraulic callipers, electric brakes would require powered actuators to move the callipers, this requires power. You can’t deny steer & brake by wire require electricity that a mechanical connection doesn’t. The question is how much? Does it cost you 2 miles or 100? I genuinely think …..by wire is a better solution & the right way to go but it comes with a hit to the battery range.
@@thelimitingfactor Perhaps, but I’d suspect it might be more than we might suspect, just the steering actually requires multiple motors that actually need a pretty strong force, with precision that operate every second the car is moving. I doubt it would be negligible. Same with the brakes. Surely it’s gotta be at least a 20 mile hit. Hopefully someone actually does the research as opposed to assumptions.
All these "by wire" actuaters introduce lag ... so you want to add laggy brakes to a laggy steering? maybe for the robotaxis but I wouldn't want that in my car for sure.
Why is Tesla going with a hybrid process to showcase on 10/10? What are the challenges of bringing the unbox process if they have all the components you describe in the video?
They're doing it to maximise the output of the current lines Fully optimized new lines with a lot of capital outlay or much less capital outlay with a process that's still more optimised than any other automaker is using?
Current production lines are well below capacity Hybrid process will allow the re-purposing of some of the existing lines to build new vehicles using non-optimized unboxed process, so we can get product to market faster Use of Optimus may make some of these non-optimized processes moot
10/10 is about showcasing Robotaxi. Tesla have said that Robotaxi will use the unboxed process (not a hybrid of current and next gen processes). Elon mentioned that they will introduce another couple of things on 10/10, but we'll have to wait and see what they turn out to be.
Now, when you have a fault in a sensor, you never know if it's the sensor or the wiring. It would seem to me that the ether loop would simplify diagnosis and repair of sensor faults.
If a light (or any unit) fails, the system will receive feedback from the related controller. If the "ring" fails at one point, there's always a connection to the "output pin" at the main computer. Then, insert 1 line within the loop from an output to an input on the computer, and you have what's essentially a "fuse" to detect "ring broken" while all systems remain operational.
Jordan is an absolute treasure. He simplifies complex technologies for us to grasp. Tesla is so far ahead it is hard to truly understand the revolutionary developments without Jordan’s work. Thank you.
❤
IMO, one point of steer by wire that is largely overlooked is the fact that still large parts of the worldwide market is righthand driving. Common parts and processes can be greatly increased between left hand and right hand models - ifyou have electronic steering, breaking and accelleration.
🎯
@@thelimitingfactor
The steering/ pedal box can be essentially a "Game controller"
("Tesla, controlled by Logitech"? 😉)
.
Great for a common Compact/ Taxi design.
Build the Compact first, *Lease* as a driven cab.
Wait for regulation.
.
When approved, THAT NIGHT, delete the modular control box from the line, change firmware and you're building CyberCABs.
.
Take the stock of steering modules to the next jurisdiction (Europe?) and repeat.
One other advantage of steer-by-wire and brake-by-wire is that you can have a vehicle that has a wheel and pedals, but when in robotaxi mode, it can disable the controls, rendering it immune to interference from passengers
(and it still has the controls for manual operation, either when just in manual mode, or when the software is unable to handle it)
Also, would probably be harder to cut someone’s ‘brake lines’.. especially without the main computer knowing before you ever start
Every piece of technology they've been developing is falling in place so beautifully!
Amen
This is incredibly concise, a bit like the process itself!
damn, you always blow my mind man. Thank you. Proud patreon for 2 years or so now
Thanks for the support! It makes the channel possible ✊🏼🤠
The quality of this channel is immense! Nice visuals, nice voice, a clear structure *and* high quality information condensed. Bravo!
Thanks man! 🤠
This info is really valuable. It encompasses Tesla car engineering, what they have already done and what they might do. Great work!
My company MAEC (modular affordable efficient car) invented the modular/unboxed process in 2007. I just didn’t have the financial backing to make it happen. Electrification is the real key that allows simple interfaces.
I love anecdotes like this!
Thanks for your input
I invented it before that, in 2005
@@daveinpublic Willy’s did it in 1940s with their jeeps. #Unboxed
😂
Super exciting that we get to follow this developmen tin realtime.
The final product is truly the future.
Thanks for the video!
Your dedication to be precise and cover benefits as well as caveats is appreciated.
Easily recommendable to other curious parties =)
Sure thing man! 🤠
thanks for watching!
Steer by wire makes it easier to make cars for both left hand and right hand drive markets as the steering mechanism is the same for both types. The only real physical difference between the two configurations is swapping the locations of the steering wheel, brake and accelerator with the glovebox in cars like the model 3 and Y.
🎯
Those items are easily modularized and you can probably just have 2 different plastic knee bolsters
Just hearing Musk saying parts of the unboxed process will be used on current production lines is exciting. Seeing this tech in use is when theory moves into being real.
I agree that if the system is fully developed, it will be a huge change for industry.
Keep up the great work Jordan!
Something I’d like answered. 12v systems, even with larger wires, have disastrous corrosive reactions in high-salt environments-wires and connectors literally falling apart. Won’t a 48v system with thinner wires and larger voltage differential have an even more likely corrosive failure potential?
It’s like building rifles; upper receiver, lower receiver, … it enables a whole secondary market to tweaking and providing modules. The legacy car manufacturers were the ones integrating many sub modules from suppliers; Tesla adds an extra layer of modularity, enabling more to be parallelized and iterated separately. It will increase development iteration cycles inside modules 🤯
Nice analogy!
I like how although there are many more redundant parts that there is the potential for the car to specifically notify you of exactly what is wrong when something does fail so the repair can be very exact.
I'm here just for the calming voice.. ASMR
I’m here for the UNBOXED PROCESS
Unboxed process sounds like the way to make everything easier while reducing costs and complexity. They could even use it to increase production for all the other vehicles (S3XY) by adding an unboxed production line for those. Also, knowing that Optimus will start working at the factory soon, I can imagine pretty much all the work could eventually be done by robots. Elon's dream of a gigafactory that looks like an alien dreadnaught will be fulfilled. But what I'm curious about is the thixomolded magnesium gigacasting you talked about in former videos. I am hoping it would be introduced alongside the unboxed process as it would imply another important improvement in manufacturing by reducing the weight of the cars!
The biggest issue with magnesium is the low ductility in crash scenarios, very problematic
@@seancollins9745 I thought the alloy that was being developed solved all those problems. And anyways, even in the worst case scenario, wouldn't it be possible to make the Gigacasting of the battery pack of Magnesium to reduce the weight but keep making the front and rear gigacastings of aluminum to keep those crumple zones the way they are, but that would be only if making them of magnesium didn't work, which I doubt.
Awesome video as always!
Glad you enjoyed it! 🤠
@@thelimitingfactorI too normally rewatch your videos 3-4 times over a couple of days as they are jam packed with incredible insights and very detailed. I wonder if Tesla will go exploring magnesium Giga casts over time? That’s a different kettle of fish!
I'll be glad when production begins.
Thank you Jordan!
Thank you!
I love your videos!! Please make more…
As usual brilliant and simple to understand. If that's your real voice btw, it perfect for what you do on these videos. 10/10
Yep! That's my real voice 🤠✊🏼
@@thelimitingfactor It's ironic. The quality of your narration is particularly consistent. Computer generated voice is not as good as high quality human narration as yours is, but computer generated voice *is* consistent, bordering on repetitive/monotonic. Nowadays people take that consistency as a cue that a narration might be computer generated. You can't win.
I believe it will help repair by allowing to take totaled car and reuse sections for a new Frankenstein car. I’m also thinking maybe they find a way to share fronts between models; model X, Y, Van, and Non-Cyber-Truck. 3, robocab, and maybe a cabriolet. So there could be some system of LEGO like fronts; heavy, high, undercab, narrow, big frunk etc. then different side profiles, different rears, and different top+bottoms.
Brakes don't necessarily need to be electric, at least in the front. Just like the steering wheel, it could be fully contained in the front module.
The rear, on the other hand, wouldn't make sense, as mentioned. But rear brakes are smaller than front brakes, so another option may be to eliminate entirely, and just brake regeneratively (especially in the Robotaxi). "The best part is no part" philosophy makes this worth consideration.
💯
VW uses rear drum brakes in addition if the goal is to maximize motor braking a resistor mounted to a heatsink could be used as a place to dump the remaining energy. It would then be vented out of the car by a cooling fan.
BYD is likely watching this video. 😎
This is an amazing analysis. All of these upgrades could have been developed by the old automakers but they just refuse to innovate...
If more companies were like Tesla humanity would be already in the age of abundance
But Tesla & Elon are unique ❤
forgot to mention the octavalve for cooling
Not necessary for the unboxed process. Certainly helps simplify things though
2 advantages.
Castings.
The ability to machine dimensionally accurate mounting faces for extremely precise location in X,Y, and Z axes.
No accumulated tolerances.
.
Etherloop.
Main(?) safety advantage?
.
Its "a loop".
A single, but complete break in that loop still leaves every individual module a connection to "home".
If you have a controller at "12 o'clock", and modules at 4+8 o'clock
A break anywhere in the ring still leaves at least one clockwise or anticlockwise route to "12".
A single test wire in the loom connected to different pins could detect a problem and advise service required, but not disable the systems.
The 10kt gigapress is big enough for door ring assemblies, let that sink in😂
The hydraulic brake lines could just use unions, many linesets on cars already do come and are assembled in multiple sections, usually on unibody cars there are lines on the subframes and lines on the body side, i don't think electric brakes are a requirement for unboxed construction
Thanks!
You're most welcome!
Great video! Learned so much, thanks!
I have a comment to the Tesla Etherloop wiring video. I wanted to submit this comment to that video of course, but I was late, and by that time you were no longer reading the comments to that video, by the looks of it. (No criticism there, you have to stop at some point.)
In cybersecurity there is the concept of "air gapped" A mission critical system should not have network connection to other components in the same facility. The obvious example is a nuclear plant. The offices that handle logistics must have internet connection, but for security reasons the computers running the nuclear reactor must under no circumstance have network connection to the rest of the facility. If the office computers are hacked, that is bad enough. But as a bare minimum the nuclear reactor controllers must operarate on an island, no network connection to the rest of the facility.
The reason for keeping the nuclear reactor controllers on an island: if a target is very attractive for an adversary then the adversary will allocate resources towards hacking that target in proportion to the possible wins. The thing is: try as you might, it is impossible to write unhackable software. There are diminishing returns. Even if you set up tons of quality controls, that make the software 10 times or 100 times more expensive in manhours to develop, you can still only make the amount of errors smaller, not zero.
It is extremely irresponsible to go with a design of cars where life critical systems and entertainment functionality are sharing a network. Promises that the critical system communciations will be encrypted don't cut it. Try as you might: it is impossible to write unhackable software.
Mitigating efforts:
Like many other companies Tesla offers large financial rewards to white hat security experts who develop and demonstrate security breaches. These experts do succeed in finding ways to get high privilege access to Tesla systems. It is to be assumed that enemy nations are in the process of stockpiling ways of access that they do not disclose to anybody else. (Such stockpiles are always in flux, vulnerabilities that are discovered and reported are patched, so new ways of access must be found and added to the stockpile.)
When white hat security experts find a way it invariable starts with gaining high privilege access in the entertainment functionality. That then forms the platform to subsequently find a way to access and gain high privilege access in the life critical systems.
A fleet of cars with shared critical systems and entertainment system is for an enemy nation a sitting duck.
About over-the-air updates
Tesla has a policy of over-the-air updates.
That policy puts a huge responsibility on the security of development and distribution.
For comparison: in 2017 the utility software ccleaner suffered a hack of their distribution servers that went unnoticed for weeks. For weeks a particular ransomware was distributed with every update of the ccleaner software.
If the Tesla distribution servers are ever hacked then in minutes all of the Tesla cars can be infected with malware.
When the distribution of updates is not over-the-air then the update cycle is a much slower process. The only practical option is that updating the software of the car is performed when the car is brought to a shop for periodic maintenance. Slower distribution is beneficial for security: better probability that if somewhere in the custody chain the software is compromised the tampering is discovered when the number of cars infected is still comparitively small.
Jordan, I hope I can persuade you to allocate time to research these matters, and to communicate to your audience what they need to know to make an informed decision. The reputation of your channel is that you are actually researching your subjects, rather than repeating the content of press releases.
Good points! I'll add it to my (extensive) list of potential videos. As I've said in the past, there needs to be other channels doing what I do.
If the system is brake by wire why are there hydrology components in the Sensify system?
... just wondering... whether brake by wire will be part of the un-boxed process... ( ie. replacing hydrolic lines ) ?
as long as the system is robustly tested I am not too concerned with electric brakes. There is no redundancy in hydraulics if you spring a leak all your brakes stop working. With the ther loop you'd need two lines to loose communication before you have a failure already making it safer. Also the duty cycle of brakes is much lower on electric cars which should increase reliability over a given time because of the lower use
Excellent points!
"There is no redundancy in hydraulics if you spring a leak all your brakes stop working. With the ther loop you'd need two lines to loose communication before you have a failure already making it safer. "
There's redundancy with hydraulic brakes. You would have to puncture both circuits in order for that to happen.
@yourcrazybear yep. Dual circuit brakes have been a thing since the late 1960s.
Excellent video. Particularly the discussion of electric braking. For almost all "normal" driving (e.g., not "off-road"), computer control of braking (for regen or for anti-lock or perhaps for skids or wheel-slipping) offers much greater safety and fewer parts, etc.
Hi Jordan. I haven't heard it mentioned yet, but in-case brake-by-wire couldn't get approved, it's still possible with hydraulic brakes by using quick disconnects. There are refrigerant QDs to disable part of an HVAC system, even though a very small negligible amount of air is introduced upon connection. There is no reason that a hydraulic QD couldn't work as well. Each sub-assembly could be either primed or under vacuum to accept the fluid when everything is connected.
This may be a pet peeve, but I don't like the term "reduce complexity" - a lot of these decisions move the complexity earlier in the process or inside a part. Internalizing the complexity DOES aid in final assembly but it does not get rid of it.
Well put
Where there is a brake on each tire, the brakes are already redundant. Each individual tire would not need redundancy.
This ~14 year old software company is showing the old timers how to build cars
We've heard the analogies about aircraft having fly by wire systems for decades, so we needn't worry about cars having the same system. Another aspect is the sheer size of these aircraft, whether they be a fighter jet or airliner. The forces needed to move the controls are far beyond what a human can generate, and as much as 200 feet from him. You can safely drive a car if the brake power booster or steering power assist fail, but you're not going to be able to fly a 747 by hand. Brake by wire will work just fine, as it already does with electric controls on aircraft. Hydraulic lines merely transfer the boost force from the power assist over the distance to the brakes, where with electric systems the boost assist, commanded by wiring, is attached to the brake disc or steering rack. You can easily have pairs of motors, and even if they're used in tandem, the loss of one merely means the loss of half your boost, not all of it.
F1 cars use brake by wire systems, you know the fastest accelerating, hardest cornering road vehicle's made by mankind.
The now 200+mph capable formula e gen 3 cars this season eliminated the entire rear friction braking systems relying instead on brake by wire all around and the rear motor itself for all of it's braking done through regen only!
Also throttle bodies and steering systems have been by wire for decades now to varying degrees. 😀
I’ve been waiting for people to dive into the unboxed process. Thanks for covering it! Joe Justice did a great presentation on this in Paris last year. Besides his videos, there’s not much information about this crucial part.
Especially when you consider the potential of Optimus assembling the modules. If we can get the module components down to weights Optimus can handle (currently 20kg), it’ll be much more cost-effective to retrain an Optimus compared to using costly, static robotic arms. This is likely what Elon means by Gigafactories being “the machine that builds the machines.”
Love your content! Check out Joe’s presentation if you haven’t seen it: ruclips.net/video/ah2O76h-c_Q/видео.html
I hadn't considered the advantages for unboxed with electric brakes. Also, question. Previously, Tesla had a patent for "rigid" wiring as an aid to assembly. I can't see any reason the ribbon wiring can't be rigid, do you think they will introduce that change as well in the unboxed process?
I think it comes down to use case. In some places rigid makes sense and others flexible might make sense.
For main power lines in the high voltage system, it's already rigid
I'm no electrical engineer, but couldn't the engines also be used as active brakes? So not just regenerative braking, but reversing the current when needed? Also, when adding a capacitor to store some of the energy might reduce the load on the battery pack, while allowing this regenerative or even active braking when the connection to the main power is lost.
And could such a setup not completely eliminate the need for disk brakes?
interesting, can we can an electrical expert to comment on this
Someone suggested using the motor case as a resistor. I'm sure there's some sort of creative solution as you're suggesting. But, that's beyond me! Electronics cook my brain
Regenerative breaking is reversing the current. Nothing more you can do there. Sorry
@@jan-ludwig5216 Thanks for the explanation :-)
I'm no electrical engineer either, but my best guess is that there are possibilities. When the motor is used for braking then kinetic energy of the car is converted to electric energy. If it was just up to the motor: I think that the motor can deliver the same braking performance as acceleration performance. The electric energy that is generated during hard motor-braking has to go somewhere. The limiting factor is ability of the battery pack to absorb that energy safely. It could be, I don't know, that an attempt to charge the battery pack faster than it can absorb will result in runaway overheating, and then the whole car will burn.
But what if you had an additional pathway for dumping energy?
When the model 3 was introduced it was described that the temperature management system has the following capability. In cold temperature the motor can be controlled to run differently, in a way that generates a lot of heat in comparison to the amount ot propulsion that is generated. Usually you want to minimize the amount of heat that is generated, both during acceleration and during regenerative braking, since heat is wasted energy. But in the case of an emergency stop it would actually be beneficial if the motor generates a lot of heat and comparitively little electric energy. Then quite possibly the rate of production of electric energy produced is within the ability of the battery pack to accept incoming power, with the rest of the energy going to heating up the motor.
After the emergency stop the temperature management system will have to start work to bring the temperature of the motor back down again. Electric cars can do that: the motor does have coolant loops. The motor has high efficiency, I think something like 97 % of the electric energy is converted to propulsion, but at highway speed that 3 % does add up, which is why the motor does have coolant loops.
Re: Your comments on braking. The motors can create reverse torque in an emergency. I think they already do! I had my car initiate emergency braking when a deer came into the road. I was already on the brakes fully (or nearly fully) when the emergency system kicked in. I'm ~99% sure the brakes weren't getting the job done and the car activated the main drive motor in reverse.
Remember when Munro & Associates dubbed entombing 4680 batteries the Pink Foam of Death? 🤣🤣🤣
💯
Nice explanation. Rather than routing the ethernet and power cable after the assemblies are joined together, couldn't they have connectors at the margins of each assembly and then simply clip them together when they are joined? Another question: I presume, as you mentioned, the assemblies will be bolted together rather than welded because the welding would damage the post interior finishes and paintwork. How will bolting compare with welding for strength?
Presumably after most accidents, the car may be partially "unboxed", the damaged unit replaced, and then "re-boxed"? Or maybe a sub-assembly could simply be removed and replaced directly?
I wonder to what extent the lifecycle emissions associated with each Tesla vehicle will be lowered by the unboxed manufacturing process, in-house manufacturing of 4680 cells, vertical integration of lithium refining, and the replacement of all ICE vehicles used by Tesla and their suppliers with Tesla manufactured vehicles. Currently, it takes three years for the additional carbon required by a Tesla in manufacturing (compared to equivalent ICE vehicle) to be offset during use of vehicle. When does that become zero?
I love that question
I'm sure Tesla said it was less than 3 years (improving!)
.
Vertical integration, if you can get it into 1 factory, reduces component transport cost/ emissions.
Then, use the SAME components across ranges to compound that.
.
EVENTUALLY (the BIG one)
.
STOP building new cabs.
Return them to what USED to be a "Terafactory" for *Refurbishment*
Remove parts, units, assemblies and rebuild those inside the factory, fit previously recycled units to the body/ battery (50 years?) and send a "remanufactured" cab out the door.
.
Reprocess the waste as raw material on site.
Who said 3 years again? I haven't seen anything peer reviewed claiming anything like that long recently done. 🤔
The latest I've seen show evs as paying the additional emissions back in 8,000-18,000 miles depending on the class of vehicle discussed, pack size, location of the owner, grid mix emissions and who manufactured it.
The average tesla sold is under 11,000 miles in the U.S. based on our average grid emissions for break even. Less if you have your own solar panels or a high mix of nuclear&renewables available. Plus the grid is greening itself month over month&yoy at break neck speeds! 😎
As far as I remember it was three months (or six). It was not years for sure as the energy consumption for the production is already only marginally higher than that of a comparable ICE car.
There are at least 4 major ways to get to zero additional carbon
1) Increase the operational life of the vehicle (use long lasting materials like stainless)
2) Increase the operational duty cycle of the vehicle from 5% to 50% (roboTaxi)
3) Make the cars rapidly refurbishible. (replaceable batteries, cushions, etc)
4) Reduce the likelihood of loss of vehicle in a crash (FSD)
Thanks
Sure thing man!
Picture of the robotaxi design looks like Tesla thought of making it a one row 3 seater option.
I'm not sure the drive by wire feature has any cost benefits for smaller vehicles and wonder if that will be left out of most future vehicles. The 48V Ethernet archictecture seems like an extermely important evoluation for cost, reliability, and manufacturing ease. I realize you can't have drive by wire without the 48V architecture, but you pointed out that the cheaper steering rack can still be used with unboxing process. Mabye the new Roadster needs drive by wire for better control at high speeds, but will Model Y and 3 sized vehicles need it?
Autonomous driving does better with steer by wire. At high speeds, you don't even need power steering in a conventional car.
@gordontyler9317
48v
@@rogerstarkey5390 changed, thanks
@@jamesengland7461 so the vehicle uses less energy? That might offset the extra hardware cost if it has longer range
@@gordontyler9317You are forgetting a major point. Manufacturing Scale. Just like chip makers turn off functionality to get a new skew and therefore have just one part to manufacture... Tesla can quite easily have every single model use the same Steer-by-Wire unit. Only the rack would change for different geometries and sizes. You can easily save enough to put Steer-by-Wire on every car.
good job
Could Tesla use capacitors so that generative breaking could shoulder more power and create longer range?
I think the electric brakes could rely almost entirely on using the electric motors. Maximize resistance during regenerative braking to the point of stopping the motor.
Always puzzled me why generative (as opposed to regenerative) brakes have to be limited by the capacity of the battery to absorb the energy. Just put an electric heater in the thermal control loop and dump any electricity the battery cannot absorb into that. This is what friction brakes do anyway, turn kinetic energy into heat, but this way there is no brake wear and sometimes the system will be able to put the heat to good use for heating the cabin or preconditioning the battery.
Hydraulic quick connect unions would also be compatible with the unboxed manufacturing system and cost way less than motors on the wheels as hydraulic calipers are small, lightweight and reduce unsprung weight as oppose to having motors onto the wheels. The unsprung weight has major effects on vehicle performance. There is no valid reason to eliminate hydraulic braking systems when making the the unboxed car. The plumbing would simply be broken up into sections and joined with quick connections, then the completed hydraulic system would be bled just like any other vehicle!
Don’t overthink it Jordan! There are ramifications you aren’t considering, when it comes to braking systems.
Hi Jordan, do you know why Tesla did not go to brake by wire for the Cybertruck?
Because it's bigger and needs something more powerful
@@thelimitingfactorHi Jordan. In your recent podcast for Farzad, you mentioned that Tesla said at their investor conference that the affordable Tesla wouldn't use the whole Unboxed process, but here Cybertruck uses 6 out of 7 pillars of it, and you said that Model 2 could use brake by wire because it's lighter. So what's stopping them from going all in on Unboxed Process could you explain?
Great video. Jordan I am wondering if Elon or someone from Tesla sees your videos? Did you get any feedback from them?
They're not allowed to give feedback, but they do watch them
The leader of their 4680 program did make a comment on X saying I was the most accurate source of information
Hey Jordan! What do you think about CATL's condensed battery with 500 Wh/kg energy density. You should do a video about it!
Too expensive 😉
@@thelimitingfactor wow you replied. Why are they expensive? I would like a deep dive into that topic.
Would be cool to see a video that talks about the impact that steer & brake by wire have on battery & total range. Are they a significant compromise?
Brake by wire?
MAY be an offshore of regen, in which case, positive impact?
Steer by wire?
Tesla already uses dual steering servo motors capable of FSD, so NO input from the driver.... So no (extra) impact
.
And the power requirement may be lower
@@rogerstarkey5390 For Cybertruck steering, every little bit of driver input, every little micro movement has to be driven by the steering motors, they take a lot of power as steer by wire was only possible by 48V. Maybe that’s partly why range of the Cybertruck was woefully disappointing. Tesla vehicles already use regen braking but they still require mechanical braking from the hydraulic callipers, electric brakes would require powered actuators to move the callipers, this requires power. You can’t deny steer & brake by wire require electricity that a mechanical connection doesn’t. The question is how much? Does it cost you 2 miles or 100? I genuinely think …..by wire is a better solution & the right way to go but it comes with a hit to the battery range.
Minimal. The amount of power used for those systems is negligible compared to the amount used to move the vehicle down the road
@@thelimitingfactor Perhaps, but I’d suspect it might be more than we might suspect, just the steering actually requires multiple motors that actually need a pretty strong force, with precision that operate every second the car is moving. I doubt it would be negligible. Same with the brakes. Surely it’s gotta be at least a 20 mile hit. Hopefully someone actually does the research as opposed to assumptions.
I think i have seen somewhere that Brembo is working on electric brakes with Tesla
Brembo is a key supplier, and brembo is working on electric brakes, so doubtlessly they've discussed it
👍👍
All these "by wire" actuaters introduce lag ... so you want to add laggy brakes to a laggy steering? maybe for the robotaxis but I wouldn't want that in my car for sure.
Why is Tesla going with a hybrid process to showcase on 10/10? What are the challenges of bringing the unbox process if they have all the components you describe in the video?
Could be risk management, or it turned out to be faster implemented if they start with a hybrid solution.
They're doing it to maximise the output of the current lines
Fully optimized new lines with a lot of capital outlay or much less capital outlay with a process that's still more optimised than any other automaker is using?
Current production lines are well below capacity
Hybrid process will allow the re-purposing of some of the existing lines to build new vehicles using non-optimized unboxed process, so we can get product to market faster
Use of Optimus may make some of these non-optimized processes moot
10/10 is about showcasing Robotaxi. Tesla have said that Robotaxi will use the unboxed process (not a hybrid of current and next gen processes). Elon mentioned that they will introduce another couple of things on 10/10, but we'll have to wait and see what they turn out to be.
Design Needs Emergency Steering & Brake in Case of Road Blockage from Breakdown/ Disabled Vehicle etc.
if you’ve ridden a bike you know you can lturn ike the cybertruck just not as quickly or maybe swiftly 🎉
😊👌👍
Most electronics will be bigger but, motors, and cables smaller
Now, when you have a fault in a sensor, you never know if it's the sensor or the wiring. It would seem to me that the ether loop would simplify diagnosis and repair of sensor faults.
✊🏼
If a light (or any unit) fails, the system will receive feedback from the related controller.
If the "ring" fails at one point, there's always a connection to the "output pin" at the main computer.
Then, insert 1 line within the loop from an output to an input on the computer, and you have what's essentially a "fuse" to detect "ring broken" while all systems remain operational.
Oh getting rid of hydraulic brakes would be so good! Screw the hundreds of collective years spent pumping the breaks when changing the stupid fluid
Amen!