Pragmatic 83: Tesla Part Two

3 November, 2017

CURRENT

Elon Musk is considered by many to be a technological visionary. In this, the third in a series of shows about Elons projects, we look at Tesla, its history, its cars the practicalities of what Elon’s trying to achieve.

Transcript available
This is part two of the episodes regarding Tesla with my guest Caleb Elstom from the Tesla show and Part one was out a few weeks ago And this is the continuation and conclusion of the episode all about Tesla. Welcome to pragmatic Pragmatic is a discussion show contemplating the practical application of technology. By exploring the real-world trade-offs, we look at how great ideas are transformed into products and services that can change our lives. Nothing is as simple as it seems. Pragmatic is part of the Engineered Network. To support our shows, including this one, head over to our Patreon page and for other great shows visit engineer.network today. I'm your host John Chidjy and today I'm joined by a special guest Mr. Caleb Elston. How you doing Caleb? Hey John, I'm doing well how about yourself? Very good, very good thank you. And this is the continuation. What's it what's gonna be exciting is to see how this unfolds in the next five years and watch all the other manufacturers try to madly scramble. I'm which I'll talk a little bit about in a minute but I'd like to guess I guess I want to keep keep moving which sort of talking about the vehicles and the branding, but one of the biggest parts for me is the battery technology in all these cars, because electric cars actually were invented a long time ago, prior to internal combustion engine vehicles. Most people don't realize this, but the truth was that the batteries were never, the power to weight was never right. And for, because the lead acid batteries that's been around since the 1800s, they just didn't work that is too big, too heavy and too dangerous, especially the wet cells way back. The lithium ions that they're using or the lithium batteries that they're using, some people have said, "Oh, well, these things can catch fire and they can be dangerous." I remember that during 2013, there were more Model Ss on the road. There was a series of accidents where there was a puncturing from underneath the vehicle, a very specific set of circumstances, but it caused a puncture of the battery pack and Tesla basically introduced armor plating underneath those battery modules to protect them from a pinhole intrusion from objects lying on the road surface. And what they've, what Tesla have done with the battery packs themselves is actually really fascinating, not just the armor plating, the whole structure of them, how they cool them. It really, really is fascinating. And the other part of it that I find interesting is that the 18650 cells in the Model S, I more commonly refer to as the 168As. They're pretty common in laptop batteries and oddly enough e-cigarettes, bigger than AA's and they're 3.7 volt cells as opposed to one and a half volt cells but they're still pretty common as mud, you know, lithium ion battery. And I think that as the acknowledgement of the importance of these batteries comes down to the construction of the the Gigafactory. So they knew, Tesla knew that part of their, their entire business was was founded on the ability to get low cost high production rate batteries and that without that they couldn't possibly succeed. The other ones charging which we'll talk about next, but the batteries that they have themselves, I still find it fascinating that they're using a standard form factor cell, but I think that they are working on other different size, different configuration batteries in future. I'm not sure for the Model 3, I know the Model S is using the 18650s? Yeah, the model 3 is using a new cell size of 2170, so it's both wider and taller, so it has less casing percentage-wise, so it is just more dense, sorry, it just has more volume relative to the case. And yeah, I think one of the big insights they had was that the production of commodity lithium ion cells for batteries inside of consumer electronics was pushing down the price and increasing production. And so when you can piggyback off of someone else's expansion, then you don't have to pay the price of the development. And so there were already all these cells being made and people like Panasonic and Samsung were building more and more of these cells for people like Apple and Dell and all these other consumer electronics, that while they weren't necessarily perfect for electric vehicles, the distinguishing factor of the price going down between 5 and 7% a year for the same energy capacity was enough to outweigh the slightly ungangly nature of chaining together 7,000 of these things into a car. I think that sort of shows a distinct difference between them and the traditional automakers who've tended to use a different size cell. They tend to use a prismatic or a pouch cell, which they only have a couple hundred of them in a vehicle. And it might be technically safer because it's slightly less likely to have thermal runaway and it's less complex. So from what the car companies care about of making it slightly easier and simpler, those cells are better, but they're much more expensive. So Tesla went this other route of saying, well, the thing that we need to eventually get to our goal of really mass market cars is cheaper batteries. We're going to pay the engineering and risk factor now of designing a battery pack that can accommodate these thousands and thousands of cells and cool them well and protect them. And I think that's, you know, now they're at a stage where they're going to be manufacturing those with their partner Panasonic, and they'll be producing so many that they can't even the big insight for making their own was like, well, if we're actually gonna produce half a million to a million of these cars, we're gonna be the largest consumer of batteries in the world. We're not gonna be able to sort of swim in the drag of all these other major electronics companies, we're gonna be the biggest one. And if we're the biggest consumer, we should also be the biggest producer because these other companies might not meet our demand. And so they decided to build their own factory and invite some of their best partners like Panasonic and others to build cells and then packs in one location to cut down on the logistics and to optimize for what the cars need versus what the consumer electronics usage needs. And so they've both designed a slightly different chemistry that they continue to refine. They've got PhDs and distinguished professors all around the world who work with them to continue to enhance and improve the trade-offs between the energy density, the specific energy, and the long-term durability, which is sort of the triangle for batteries. And then also optimizing the physical size so that they can pack as much useful material into that sled underneath the vehicle. Now obviously with fewer cells, they have less surface area to cool, so they've also had to invent new cooling pack designs. But what's interesting is in the most recent 100D and 100 packs that they're using in the Model S's and X's, they sort of brought forward some of the technology they were going to use for Model 3 to validate it at the slightly lower scale of Model S and X, and that's been going well. And so now that's sort of the general cooling infrastructure they're going to use for the Model 3. So, you know, not relying on a third party supplier like LG Chem, which is what GM has done with the Bolt, gives them the ability to actually push on this. And because it's the single biggest reason you can't have a $20,000 or $10,000 electric car that has 300 miles of range, it's not the motors, it's not the seats, it's not the sheet metal. Like obviously we figured that out with other vehicles. It's the battery pack. And so that is the number one critical path to getting these cars to be in everyone's driveway. And Tesla is trying to do this and they've built this Gigafactory in Sparks, Nevada. It's going to be the biggest building ever kind of hyperbole that Tesla loves. But it's not just for hyperbole. It's because they are going to need an immense, immense amount of capacity for batteries and moving all that mass from, you know, overseas is not smart, especially if you need, like putting batteries on a boat and exposed to seawater is not a good idea. - No. - So they're producing them in a very dry place in Nevada, which is great. And they're gonna be building more of these gigafactories that eventually are also gonna be the car factory. So they're gonna have raw materials come in one end and cars come out the other. And so they've said they're gonna announce two, three, four new gigafactories by the end of the year. And yeah, I don't think they get enough attention for this. And you see folks like Daimler and others now starting to announce they're gonna be building their own battery factories in partnership with some of these major manufacturers because you can't get enough of these cells. If you went on the open market and said, "I wanna build 100,000 cars," you may not be able to get enough batteries at a cost you're comfortable with in the open market because they're all being used and you can't just build capacity instantly as Tesla has been working on the Gigafactory since 2014. - Yeah, that's right. - So it's a long-term planning and I think one of the things that Elon Musk is pretty good at is this idea of really thinking through what are the fundamental things that need to happen for this goal to be achieved and trying to make sure that he is in control of those major attributes and making sure you have enough batteries if you wanna produce a million cars a year is pretty critical. And he's the kind of person that says, well, I guess we need to build the biggest battery factor ever. So let's go do it. - Yeah, let's do it. Exactly. I think it's wonderful, the fact that they thought it through to that level, understood their dependencies, and they didn't hedge their bets, just like Elon did with SpaceX and Tesla in terms of investment. He went all in, they've gone all in with the Gigafactory. There's nothing small about it. Gigafactory in every sense of the word. And it's truly impressive. And I'm excited to hear about the two, three, and four gigafactories as they progress forward on different continents and try and ramp that production up. In terms of the battery packs themselves, though, I found it interesting if you look at all the different manufacturers-- I know there's not a huge number at this point in terms of electric vehicles, but the LEAF that I have driven, for example, the LEAF does not have a cooling system. It's separately. It's naturally cooled. It's, and the early models of the Leaf have known battery degradation problems and you know, you lose X number of bars of capacity and you are entitled to a replacement, but the newer so-called lizard style of battery has improved cooling, but it's still not the same kind of unit. And in the end, Tesla figured this out early on. They didn't want to take the risk. And so they've done it using a glycol refrigerant system. And that's actually quite commonplace in high voltage variable speed drives that I deal with generally. And it's not a new thing. It's not a necessarily technologically advanced thing. But it's something that's been sort of like shrunken a little bit down into a compact battery module, which is kind of interesting and interesting application of it. And, and Tesla's works quite well. One of the things that I think people don't realize is that batteries get hot when you when you charge them or when you discharge them. So when you're used to consuming the power when you're charging them when you're charging them, the power going through it because of a vehicle's IR, which means you've got resistance through all of your conductors. So ultimately, resistance is nonzero, because we're not talking about superconductors. So in the end, when you pass that current through a conductor, you're going to get a voltage drop because of that resistance and that extra energy will be lost, manifest as heat and heat's bad, because you get thermal expansion of metals that creates micro deformities, you can get fissures or leaks within cells. If you don't manage it and as in you don't extract the heat out then you know you're going to have problems and ultimately you can get as you mentioned before thermal runaway. It's not a good situation so keeping those batteries cool under high load in either direction that's either discharge or charging either way is critical to get it right and I think Tesla's system is has been very well engineered and seems to be quite reliable. So and lithium ion in terms of a battery technology is still your best bet in terms of energy density for weight that can be mass-produced. There's other ones that are promising, but they're not at the point where you can mass-produce them. So, they're still the winner for now, I think. And the way that Tesla's approaching the Gigafactory is definitely the right approach. And one of these days in a future episode, I want to go through... I did an episode way back called "The Battery Problem," which is only episode two of this show. I want to start looking at because the game is changing and Tesla has been driving a lot of it which is great to see because the batteries are also now being beyond just laptop cells, they're now being used for mass storage for grid storage and things like the Powerwall for homes and the massive power packs that they're installing around different parts of the world to basically because of the surge in renewables and of course solar renewables don't work when the sun goes down. So you need to store the energy and it's quicker and easier to deploy than digging a dam for hydro storage, which has been the traditional way of doing it. And so batteries and battery technologies, they are the biggest game changer. And the funny thing is that they were driven by vehicle. And yet they're going to have this enormous impact in ripples in how we deal with energy in the world in coming decades. and that's pretty amazing, I think. - Yeah, I mean, I think that Elon recognizes that if you want a fully sustainable energy solution, he believes pretty strongly that it's gonna be mostly solar, but even if it's solar or wind or any combination of renewables, very few of those run all the time. And so you're gonna generate this energy, and then if you have no fossil fuel, you can't run with no energy at some points of the day. it's impossible for the world to work that way. So we need to buffer it somewhere. So it needs to go in a battery. And so you need both renewable generation, which is solar, solar city, now solar city is part of Tesla. You need battery storage to buffer in times when the sun isn't shining. So that's the power packs and power walls, as you mentioned. And then we need to move around. And so you can use electric energy to do that as well. And his general proposition is that every form of transportation except for rockets will be electric powered. And we can have electric heat, we can have electric refrigerators, we can have large industrial products are electric. I recently realized as well that cruise ships use electric motors for actually turning the rotors. they use diesel to generate the electricity. Yes. So it's just you know it we're not even huge cruise ships are actually using electric motors and so it we're not actually that far from being able to do this once the batteries density and cost comes down. So and I think one of the other things he talks about a lot which I think people sometimes forget is we we've benefited from such a huge advantage on microprocessor improvements but we We almost take it for granted, but hard work is being done on that every single day to make those advances possible. And batteries hadn't been getting as much love until about the 80s. So we're 20-25 years behind where microprocessors are. And we had fewer people working on them because batteries weren't getting as much attention as computing was. And now as we have more and more engineers working on batteries, I think we'll potentially be able to see the types of advances we want at maybe faster than 5 or 6% a year. But that's still a great improvement. Engines are not improving at 5 or 6%. Gasoline is not getting 5 or 6% cleaner every year. So, you end up in a good spot. It's just how quickly is the question. That's it. And depending upon which gasoline station you go to, they'll tell you that just two tankfuls of their special blend will clean out your engine magically or something and I mean I see claims like that and I just I sort of roll my eyes and I wonder one point in the future we're going to have superchargers done by a company that says that if you charge using our electrons then your batteries will last longer we'll see what the marketing spin does I guess the next 10-15 years but hey yep yeah anyway so there was something specific though about these batteries in the hop the top performance model that I really wanted to just mention and that is Inconel. And you talked about on an episode of the Tesla show and that sort of peaked my interest. So, I did a little bit of digging into this because I'd actually come across Inconel before but because I'm an electrical engineer and one of the things that we deal with is in the current industry I'm working in is some turbines and it's interesting that Tesla uses Inconel as part of the main battery pack power contactor in the performance models of the vehicles. And I started digging into this and the contactors in question for example, these ones, they're not load brake contactors by design, so they don't actually do any arc quenching or anything like that like a circuit breaker would. But the Inconel is less susceptible to scouring and carbonizing for whatever it's worth, but it's more the fact that it lets them push, I think the limit was the 1300 amp limit for the cross-sectional area on the contactor up to about 1500, so 600 to 200 amps without deforming the conductor importantly. Personally, I still wouldn't be pushing 1500 amps for that long, because it's still a very small cross-sectional area for that kind of current, even with Inconel. But the thing that's interesting about Inconel is that the metal is actually more commonly used in aviation and exhaust nozzles. They also use it in rockets for similar reasons, but steam turbines use it as well as gas compressors. It's been around as in terms of materials, it's actually been around since the 1940s, since the early jet days. And it's really painful to work with. It's not, it's generally considered a last resort metal than anything else. But there was a suggestion that was used because of Elon's work with SpaceX and cross pollinating with them. I don't know if that's more of a myth than actual truth, but most material scientists and mechanical engineers know what Inconel is. It's just, anyway, but the bottom line is the part they're using, I tried to dig up the exact part from TE Connectors, but I had trouble tracking down the exact part number. So I couldn't narrow down the exact alloy, because it's actually quite a large range of Inconel alloys. But still, as applications for Inconel go, one I hadn't come across before for that particular metal, and very, very cool, very high on the coolness scale, I think. But anyway. Yeah, and just stepping back, the whole thing with Inconel was that in the performance models, when you put your foot down on the accelerator, they're trying to pull as much amperage out of the battery pack as quickly as possible to pump it into the motor to spin it faster. And they were hitting a limit where they weren't able to draw as much out of the battery as they hoped because this connector, the contact for the transfer into the rest of the power electronics was failing. The metal was not conducting anymore. And so yeah, they switched to this Inconel material instead. And yeah, they claim that Elon knew about it and suggested it because of the time at SpaceX. Now, certainly, if they had good electrical engineers working on this, they probably would have known about it. But it, yeah, the facts remains that yes, SpaceX does use Inconel in in some of their exhaust housings and things like that. And I think the fact that any people who follow Tesla even have heard of the word Inconel because of Tesla is intriguing. That this might be one of those examples where the Apple fanboys get upset about Apple talking about retina and things like that, super high DPI and trying to brand it. But I don't, it doesn't matter. The point is that people think about this as a performance improvement and some cool space technology they brought and one of the other things they talk about a lot of some of the welding techniques they use are similar from from SpaceX and you know I think it's definitely cool to tell your friend that your car has some space-age technology in it from SpaceX and whether or not it would have independently been able to be invented at Tesla alone or any car maker the fact that Tesla is able to get that story out is is valuable to them and makes their customers feel more connected to what they're doing and it's a cool tech detail. And I think SpaceX does a really good job with this as well with all their press conferences and things. But yeah, I think the other car companies could learn something about how do you actually get people excited about some of this new technology and not some of just the legacy motor technology because many of the people who follow these car companies are gearheads care deeply about how the engines are improved, but they don't really spend as much time on their Bluetooth interface. No, but you're right. It's fascinating how the details of the metal used in a contactor in the car. Right. How many grams of that metal are in the car? Probably a few. Yeah, that's it. And it's suddenly this, it's got the cool factor and it's people are fascinated. I I mean, I'm fascinated by the fact that they did it, but right. I love the detail, but anyway, it is very cool. So like I said, very high on the coolness scale and it's all in the name of speed. And this is the other thing that, you know, some things that I've thought about, about why Tesla has this lustworthiness. And I think a lot of this comes back to car companies for the longest time from the days of the EV1, I think maybe, maybe before that was pushing this idea that EVs were slow. They just, they were an old, you know, like a real car's got a loud engine and, you know, lots of brake horsepower and brum brum and all that sort of rubbish, you know, and it's like, and that was the definition of an electric car was it's never going to be able to compete with that. And Elon, they said Tesla set out to prove that that was not correct. And most electrical engineers sort of knew already that electric power was far superior to combustion engines for a whole bunch of reasons, including torque as well as power to weight ratio. The issue was the portability of the power. So we talked about the batteries and we'll get to charging in a minute. But the problem in the past, you know, it's like the electric has been around for 100 years and the battery problem is the problem. And carrying that around is a problem. Lithium ion works. So that's talked about that. We'll move on. But the thing is that the power to weight ratio is far superior than any gasoline engine. And people say, oh, well, you'll still get better results out of IndyCar or Formula One. But then, you know, technically that's true. but at the same time you have to rebuild the engine at the end of every race for Formula 1 cars. So that's kind of cheating, unless you're going to rewind your electric motor at the end of every drag, which you don't have to do with a Model S. So you know, hmm, that's not really the point. So in the end they wanted to build a car that was very, very fast and push that forward and they just kept taking the next bite and the next lowest hanging fruit. So the first low hanging fruit was dual drive, so going from rear-wheel drive to dual, so all-wheel drive, to improve their acceleration and the next one was larger battery packs so the bigger the battery pack the more current you can draw and still have an acceptable range and obviously then the next thing that broke was we just talked about the contactor so they upgraded the contactor and they just keep on clawing it in and clawing it in and clawing it in and it's reaching that point of insane mode and then there was ludicrous mode and actually I love the fact that ludicrous mode is a reference to Spaceballs, at least I'm pretty sure it is. Yep. Yep. I'm pretty sure it is because he's teased maximum plaid will be coming. That's so good. I mean, really, that is amazing. I love that. Anyway, so because in mid-2016, the Model S actually was crowned the quickest production car in the world. I guess that's kind of debatable on the proviso that the LaFerrari, the Ferrari LaFerrari, And the 918 Spyder with, so like limited production runs, but you know, production, not production, limited production, I guess. But what's not- Yeah, currently in production. Yeah, currently in production, sure. But I mean, debate- But there's no debate at all that it's the fastest five-seat sedan in the world. They're at that point, sub 2.5 seconds, 0 to 60 time. But then it was actually in February this year, in 2017, that Motor Trend clocked the P100D this model year at 2.28 seconds, which is just absolutely mind-blowing. And there's no question that that kind of performance is fantastic advertising. So when you say, "Oh, I'm getting a Tesla," you associate that with this high performance and high acceleration. And I think that whether or not that was Tesla's intention or not to shame the gasoline engine into oblivion or not, that seems to be the way it's headed, which is fine with me. - And what's funny too, just before we move on, is that all the other car companies that are making electric cars could do this too. - True, yeah. - They could take this claim. I mean, even driving a Bolter at LEAF, the zero to 30 is quick. It feels good, but Tesla was able to stake it out. And it just sort of reminds me, I don't know if you ever saw a show called "Mad Men", but it was about advertising. And one of the times the main character has to try and brand a cigarette, and you know, not a fan of cigarettes, but the story is still interesting of, he'd ask them, "Tell us what you do to make the cigarettes so we can come up with something interesting and rattling off all the things they do. And at one point he says, "Oh, we toast the tobacco." And he says, "There it is. "Our cigarettes are toasted." And all the cigarettes are toasted, but linking onto this idea as an important bit and hitting on it, I think, certainly trying to, as you said before, fight against this idea that electric cars are slow and golf cart-like. So you swing to the totally opposite of, well, we're gonna be the quickest, nothing else can beat us. And the fact that it sort of comes for free in air quotes with electric motors is, doesn't matter. Even if everyone else could do it, Tesla has done it first and staked out that claim. And in branding, you don't wanna be the second person to try and do that because you'll be seen as copying. So it's intriguing. - Yeah, it is interesting. And I think that it's very smart. and the way that they've just taken the old sales pitch, which was never true, but it was the limitations of technology at the time meant that it was gonna be a tough sell for electric cars in terms of performance, but that game has changed, it's over, and Tesla have basically ended that debate. There is no more debate anymore. So, all right, so that's a little bit about the speed. Gotta talk about charging. So the Supercharger network, that was announced in 2012, it was only pretty recent. - Yeah, a couple months after the Model S started shipping. - Yeah, that's right, yeah, it was. And there were only six stations to start with, but I checked at the time of recording and I think you said about 900, I think it is about 900, with around about 6,000 individual charges in total, that's charging points. So that's about six and a half charges per charging station or thereabouts. The thing that annoys me as an electrical engineer is when they say supercharger, I mean, it's just a marketing buzzword, right? What they really mean is it's a DC fast charger and you know, that's cool, but DC fast chargers doesn't sound cool. So anyway, that's fine. That uses the Tesla's proprietary connector and the DC charges at up to 480 volt DC. Maximum is about 120 kilowatts per vehicle, but obviously based on the Model S, which was originally designed for X, I guess, you should reach about an 80% charge in about 40 minutes or so from flat, I think. but realistically you don't pull into a charger at flat. Well, you really hope you don't. - Yeah. - And so you would get to about 80% in about 30 minutes and that's what they kind of, you know, market is this is the, this is what you can expect. And then after that, it slows right down because of the laws of physics and all that to get to a hundred percent in about another 30 to 90 minutes I guess it depends on the size of the battery pack and how flat you were, but in any case. So the charges themselves had to be distributed well across the country and the world in order to make the car a reasonable proposition because if you were going to trickle charge them, you could do it at home depending upon if you had the high-powered wall connector or not. It would depend on whether or not you could actually charge it overnight or whether or not it would take several days to charge it based on how much current you could get out of your power outlet. And a 100 kilowatt-hour battery pack is huge, and charging that on trickle is going to take days if you plug into a standard 110 volt outlet which you know you could do you can do but geez you know you're going to be waiting a while for that thing to be fully charged so you know the idea of the superchargers was to fill the gap for people going between cities that there'd be enough range for you to charge overnight at home with a high power wall connector and that would satisfy your requirements and and that's fine but the Supercharger Network was an enormous investment and continues to be an enormous investment and they have to keep rolling it out because unlike the rest of the market, we saw what happens with gasoline. So, way back in the day when gasoline was rare, you wouldn't venture out from your home city, your vehicles would stay where they were and some people would do it. Of course, they'd fill up jerry cans full of gasoline and they'd go between cities and so on. But getting stranded and running out of gas was a common thing. So, a lot of people didn't like doing it. And then, of course, you know, interstates happened and highways and so on. And then they built gasoline stations everywhere along them so that you always had a place to stop. You didn't have any range anxiety. So, the thing with electric vehicles is that you would hope that there'd be similar infrastructure, but there's no incentive to build it. And so, yeah, no one built it. So Tesla had to build their own to make sure that, you know, that they would have, that their vehicles would be able to be serviced. And I think that it's been an absolutely critical part of their strategy. The thing that I find is frustrating, I mean, I understand from a business point of view, why is that it's not an open standard? So you can't drive a Leaf up or any other non-Tesla to a Tesla supercharger and plug it in to charge it. You can, however, go the other way around. If you've got something like a Chidamo or a CCS charging station, you can get an adapter. I know you can for Chidamo. I think you can get an adapter for Tesla to CCS and you can plug in and charge your Teslas at a non-Tesla DC fast charger, but just not the other way around. And I understand why they did it, but because it's like they're- Tesla have sunk all this money into doing all the supercharger network and destination chargers and such. They don't want people just plugging in and using it. they want it to be for the benefit of Tesla owners. And if you walked up, if you rolled up to a Tesla supercharger station and it was full of Nissan Leafs, you might just be a bit annoyed, I guess, as a Tesla owner. - Yeah, for sure. And I think that the important point you mentioned was that when the gasoline stations were built out, it wasn't the car companies that built them. It was individual entrepreneurs who, and the oil companies who decided we want to build these. And because the only option was to use gasoline or diesel, you knew that everyone who was going to come your way in a vehicle was going to need your product. And right now, the tiny, tiny percentage of people who have electric cars are not a large enough market. And it's not growing currently at a pace fast enough for many entrepreneurs to say, I want to build electrical charging place. And the other challenge is that because the speed of charging is so much slower than filling up a tank of gasoline, your throughput of vehicles per station is quite low. So your land utilization is not nearly as good as a gas station currently. And so you know if you want to put... and also the other huge thing is that most people charge up at home. So it's not as if they need you every single week. It's very possible to not use superchargers at all unless you're traveling once or twice a year maybe. And so the utilization of superchargers while currently is definitely a consternation for a lot of Tesla owners who find that there aren't enough. It's certainly much better than if you could only charge at superstation superchargers which is the case with gasoline stations. So the economics of it, of like public charging, I think is pretty rough. Especially when you're competing with electric rates that are already half to a third the price of gasoline. So you know when people go and I see some charging public charging stations and that's twice what my home utility rate is, I am frustrated, but I shouldn't be because they need to make a profit. They're not gonna charge the same rate as my home electricity that has all the advantages of scale and I'm paying for the real estate. So I think public charging is definitely gonna be a challenge and yeah, Tesla doesn't want to give all the other car companies the advantage of having not done anything because the majority of the car companies are not doing charging. the very few who are usually putting them at their dealership through forcing their dealers to add some charging 'cause they're selling some electric cars. And then you have folks like ChargePoint and others who are building these things. But I agree, I think the supercharger is huge. And for me, it's the only reason I can get a Tesla because every once in a while, I do go down to Los Angeles, which is about a five to eight hour drive. And I wanna be able to take my Tesla. Now I could get like a leaf or a bolt for around town driving and then rent a car, but that's too big of a change in behavior to seem reasonable, especially if you're buying a new vehicle, like to imagine you're gonna have to rent a car. And so I definitely believe that their idea that well, you can go anywhere, even though people realistically do it, a fractional percentage of the useful life of the vehicle is enough to get people not to buy your car. And for some reason, the large car companies are not willing to make the investment in the infrastructure. I think what you'll actually see is more what you're saying is they will band together for this more common open standard because none of them want to invest the amount you'd need to do the entire world, but they're willing to contribute to it in a smaller amount. And it will, of course, happen over time. Tesla's just tried to accelerate it for themselves because they want people to be able to do this right now. And honestly, it's pretty impressive. I was just in Bozeman, Montana, which is pretty far out there. And there was some superchargers there. No one there, but they were there, ready for people coming through Bozeman. So I think the superchargers are great and really, really important. And they started out as free for all the people who were using them. So even further down the path of just use it, and it's sort of built into the price. With the Model 3 and future Model S's and X's, you'll pay a small fee per kilowatt hour per time, depending on your municipality, but it should still be about half the price of gasoline at current prices. So, yeah, and the fact that it's DC fast charging is great and they're gonna continue to keep pushing up the limits. I think now it's up to 145, even though the cars onboard bypass can't even take that. So they'll keep pushing it as the batteries can keep accepting higher and higher C rate. So, you know, at some point, 10 minute charge would be sort of ideal and you'd really have no complaint against using this charging. Right now 30 to 40 minutes can still be a meaningful addition if you're doing a long road trip but if you're just doing a day trip it's probably not that bad. Yeah, exactly. And I think that in Australia for example the superchargers have just made it to Brisbane like I said a month ago. They opened a dealership here, they've co-located these superchargers there So you can now supercharge basically from Brisbane all the way down through Sydney to Melbourne and then across to Adelaide. And, you know, there's more opening all the time. You can't quite drive around Australia yet. I know I'm pretty sure you can in the States more or less. Yeah. More or less. But, you know, not here yet, but it's not far off. It's coming. All right, cool. So a couple more topics and then we'll wrap up real quickly. All right. So I just want to quickly talk about patents real quickly because this is an autopilot. So the patents now Tesla made a big announcement in mid 2014. I'm going to put this the text in the show notes rather than just reading it verbatim. But basically in mid 2014, Elon Musk penned a letter and said, you know, there was a wall of Tesla patents in the lobby and we've you know, they're gone in the spirit of open sourcing. You know, we're now going to basically anyone can use them. just ask basically, and we're not going to litigate, which is a very nice thing to say. And when I heard that, when I read that, I sort of thought to myself, well, what did what does open sourcing patents actually mean? But in any case, there's a link in the show notes that shows you that will have a whole list of all of them. There are hundreds of these of these patents, some of them are more useful than others. And most of them don't really as far as I can tell, based on when I looked at some of them, I just opened up a few at random that I really represent much of a direct engineering competitive advantage and other things like the way they mount a car seat, having a user configurable screen interface, wire break detection in redundant communications paths, which I found interesting because like everyone does that either directly or indirectly. So I found that to want to be a bit odd, but then like most patents, they'd say, Oh, we'll apply for a patent for that. And if we get it, we'll keep it and so on. And there'll be part of our, our, our leverage in future if we feel the need to exercise it. But, You know, the ones that were more interesting, though, were things like varying the flux versus torque for maximum efficiency, which is, you know, talking about the magnetic flux control for induction motors, which was actually really fascinating. There's another one about flux controlled motor management sounded pretty interesting. I didn't dig into that one specifically, but, you know, I would say that interesting considering my say it's interesting. That's just considering my background. But I just thought that that was interesting because Tesla started out going down the same road with patents as a lot of other companies. And I said, you know what, this doesn't make sense. And we're just going to chuck it. So have a read of that if you're interested and yeah, I just not sure if there's anything else to say really on the patents. I just wanted to make sure that I mentioned the fact that I think that was very impressive that Tesla did that. I don't know how much of an advantage it is or one way or the other, but it certainly is a nice gesture. And I know that other companies are taking advantage of that. I don't know to what extent, though. Yeah, I think that it's a I think it's useful as well for the team members working at the company to know that the work they do, no matter what happens to Tesla. will be available for others in the future regardless of whether other I think I don't know of many other manufacturers that have actually started using those patents and I think part of that is because to your point it's not really possible to make them open source it's just an agreement and at any time Tesla could change the terms of that agreement and then therefore be litigated so I'm not sure many legal teams would allow a company especially a major company to use those patents safely but I'm not a lawyer but I do think sort of philosophically it's a stand to say we do legitimately want to accelerate this transition and inventions we make we are willing to work with people to use but we haven't seen people use a supercharger proprietary connector they still apply for design patents for the design of their cars and their wheels they don't want people copying those so it is not complete it is not a fully open source sort of mentality, but I do agree it's a it is better than keeping them totally private, especially for the purposes of some of these more obvious patents that are ridiculous, but they were granted and every company that applies for patents gets some of those and it's sort of strange and slightly weird to use those as a weapon. Yeah, exactly. And Tesla, I think the message Tesla was sort of sending out is that we're not not going to be a patent troll and we want this industry to move forward and we don't want to hold it back and I think that that's admirable. I don't expect that they'll ever flip on that, at least not under the current management. But yeah, still. All right, quite possibly the most exciting topic then about Teslas is autopilot. So we absolutely got to talk about it. And just a little bit of the background first, because I know you've dug a lot more into this than I have, but I've sort of followed it a bit more tangentially. So the basics of it was the early Model S's used a chipset from a company called Mobileye called the IQ3 and they tended to push the limits of what it could do. There was an incident in Florida that did where someone died in an accident. And internally Tesla, I think at that point were already pushing to develop their own system but then decided to make a break from Mobileye and went with their own Tesla Vision system. October 2016 they fitted that using an NVIDIA drive PX2 computer on board and only recently their software update allowed the Tesla Vision or second generation of Autopilot to match the features of the first generation Autopilot. I think there might be still some discrepancies there but since then Intel acquired Mobileye for 15 billion. I'm pretty sure I was right 15 billion. Yeah. That's a lot of money. So Intel obviously saw the value in what mobile I provided. So and clearly Musk believes that autopilot functionality would be the key differentiator for automobile manufacturers in the future. And the funny thing is about a couple years ago, in 2015, I did an episode of Pragmatic. And I said that the next big shift in automotive technology would be self-driving. I got a lot of people message me afterwards quite unkindly in disbelief. but since then a lot's happened and I still think that that is the case and I think Elon thinks that's the case as well. It's not going to be easy though and I think ultimately yes. So how about just a little bit about what Autopilot, I'm yammering on about some of the tech behind it, but what would how would you describe Autopilot functionality as it stands today? Yeah I would say that Autopilot is a convenience feature that makes driving on the highway a lot more Comfortable and less stressful and it does that through two major features one is a adaptive cruise control So it controls your longitudinal Speed so when you're approaching a car, it will slow down if you've got wide open road It will keep you at a cruise control. So Traditional cruise control you have to set a speed and it doesn't have any awareness of what's happening So it will run into a car Traffic aware cruise control will use in Tesla's case radar as well as a forward-facing camera To detect that there's a vehicle in front of you and keep an appropriate amount of distance. So that controls your speed all the way down to stop and go traffic, which is really nice to move along in traffic without having to manually manipulate the accelerator and brake and Then the more advanced portion because many cars now have traffic where cruise control to varying degrees of quality and capabilities. But Tesla's works pretty much from a dead stop all the way up to, you know, 90 miles an hour. And then they have lateral control through a lane keep assist feature. And many cars do now have, well, they've got different things. And this is the problem is that many cars will claim they have lane keep or some sort of autopilot style functionality. And when they say lane keep, they mean we just keep you from exiting your lane. But that doesn't speak to anything about the quality of riding inside of a lane. And so many cars will let you drift to the edge and then sometimes drift over the lane and they'll just buzz or come in. Tesla does a really good job in most cases of keeping you dead in the center of the lane, even if the lane markings are slightly weak or missing for stretches. But it is sort of this, but those two things combined pretty powerful that when you get on the highway and you double-click the left stalk and engage autopilot in you you basically can Rest your hands on the wheel and not really do anything and it will cruise and having used it for a couple hundred miles Combined now in different drives. I've done with Tesla's it is at first quite strange. It feels like a ghost is driving and And then pretty quickly you get very comfortable with it and it does remove a lot of the menial work that you're doing when you're driving on the highway and frees you up to do much more of the higher level type tasks of observing your blind spots, looking around, planning when you need to move over to other lanes. And now autopilot as well will allow you to change lanes automatically. So you indicate to change lanes and it will move you to the next lane if it's clear after you've checked. But autopilot as it stands today is an assist. Every time you engage it, it tells you you need to be in full attention and paying attention and you are in control of the vehicle. And I think many people are under the wrong impression that autopilot is self-driving. Autopilot is a assist feature in a very limited domain of highway driving. Some people do use it on surface streets. It's really not designed for that, but it is the base functionality that eventually will be used to get Tesla to their ultimate goal of complete self-driving where you can fall asleep in the car and it can take you anywhere you want to go. But the time and the amount of engineering to get from where we are today to that is where people will have massive debates and fights and there's billions of dollars being invested in different companies to get us there faster. But autopilot today can cover a large amount of most people's driving if you have to use the highways and be very nice. Tesla's done probably the best job out of all the manufacturers of branding this set of functionality into a product called Autopilot that is integrated in Tesla controls and to your point they were working with a provider named Mobileye which probably was more uncomfortable to them because it had so much more software that Mobileye was doing and they weren't able to push as much as they wanted. They were able to get firmware from their hardware vendor, Bosch, for their radar and ultrasonics so they could fuse that themselves and write their own control software. But the cameras are really where Elon and the team believe that you have to solve the vision problem. You have to be able to perceive your surroundings and that they believe vision cameras are the the way to do that versus some other approaches like LiDAR and thermal cameras and things like that. So they've tried to go head deep into that and the first autopilot cars in the Model S and X were one camera facing forward and a forward-facing radar and then the ultrasonics for parking and such. And then in October as you mentioned they made a very big shift to an entire new Suite, which had eight cameras, a much better radar, better GPS, better GPU as well, better IMU for inertial motion unit measuring, and then also better ultrasonics. And so now they believe they will be able to achieve what is considered level four and level five. And so there's all these different levels, but essentially level four would mean under Under a certain domain, like a particular city or particular highway or particular space and a particular weather and time, the car will do everything and you don't have to intervene. Level five would mean there's no restrictions, it could do it everywhere. Level five is the furthest away, obviously, because you can't put any constraints on its usage. So I think the functional thing to really care about now is level four. what levels, under what sort of constraints, can you totally not engage with the vehicle? And Tesla is pushing very hard with their current sensor suite to get to that point. Unfortunately, it's taken them a while to catch up to where they were with Mobileye. So it's taken them about a year to get back to parity. And some current owners would argue that they're not at parity with the way that Autopilot One, as it's called, performs. And so when they made the announcement of the new hardware, they had all this discussion of the full self-driving. They did a demo video of one of their Model Xs driving around the headquarters doing full self-driving, which was sort of a choreographed stage demo. But it was working. They did 400 miles before they recorded the actual test that they used for their video. And then the enhanced autopilot is this idea of, well, now the car will exit the lane, lanes without you actually clicking on the turn signal. It will exit the highway for you and all these sort of more enhanced versions of autopilot. Honestly none of those have really come to bear yet but Elon has said they plan to do a cross-country trip at the end of the year or very soon after from Los Angeles to New York with no one touching the controls. He reaffirmed that just a couple months back. So we'll see. They've had a lot of challenges making as much progress as they want publicly, but the big challenge for them is obviously that they have these cars in all the major countries of the world and they need to make sure that it actually works in practice because they don't have the ability to show a demo in Las Vegas at CES on a closed course and really sell that. They're selling the real product. And I think that's been a tough challenge for them is I would expect they've got the 85, 90% done of many of these features internally, but that last 5 to 10% validation of getting it perfect or as close to perfect as possible for the level of safety they need is extremely difficult. And I expect they will eventually get there. But I'm very excited about it. But I just think that it's unrealistic to expect that in the next couple months, all the Model 3s will be self-driving. And for anyone who tells you that's going to be possible, I think they're quite... Optimistic. Ill-informed. Yeah, ill-informed of the actual challenges here. Yes, insanely optimistic, in fact. Yeah, that even folks like Waymo and Cruise at GM are working very hard. They have more sensors. They've been working on this in a very limited domain of just a few cities. and it's still not quite ready for passengers, tells me that Tesla still has a long way to go. Now that's not to say that Tesla, I think highway is a great place. I think making highway driving completely automatic where you can really just take your hands off the wheel will be incredible. That will feel like a self-driving car when you're on the highway. And getting to and from the highway for many people is not the annoying part of the drive, but the long haul, boring highway stop and go traffic is not exciting driving. And so I think Tesla's approach of piecemeal productizing particular parts of the driving aspect is important. And sorry, one other piece is they've got sort of the highway, if you think of highway as one end of the spectrum, you've got parking at the other end of the spectrum, right? And they have some functionality around summoning your car from a parking spot, parking itself without you inside of it, sort of like a little remote control car, it'll sort of park itself and come back out. and eventually it will come find you like the Batmobile would, which I think will be really fun. And so the important challenge will be connecting the highway driving, which is, I think most experts would agree is the easiest of driving tasks that a car has to perform because there's so many more constraints on the type of movement and the type of pedestrians and others. You just really don't have to deal with those on major freeways and major highways. And then this parking environment where it's very low speed, you've got ultrasonic sensors to make sure you don't hit things, that connecting those two ends of the spectrum, the city driving will be extremely difficult and doing that to a good enough degree in cities will be incredibly difficult to do. But I think they will do it and I think they're making enough investment in it. And I think Tesla and Elon understand that this is an existential important thing that if you want to be a transportation company the future you need to own and control your self-driving software otherwise you will be a supplier essentially. Yeah, exactly right. And what's interesting to me is the different timelines. And you mentioned that Elon Musk said that the end of this year there'd be a cross country trip in the States from coast to coast, you know, autonomously. And the thing is that I know that the end of 2017, that was the target and recently reaffirmed, as you said, but compare and contrast that with what Intel said post acquisition of Mobileye saying that full level 5 won't be possible until 2024. And it's interesting that, yeah, I don't know. I have my doubts. And I think that, actually, just real quickly, we should just touch on the five levels and what they are really quickly. The NHTSA, or National Highway Traffic and Safety Administration in the States. They adopted the set of suggested Automotive Engineer Society SAE levels for automatic driving systems from level 0 to 5. Real quick, 0 is the human driver controls everything so nothing's autonomous. So I have level 0. Good. 1 is a specific function like steering or braking can be done automatically. 2 is both steering and acceleration/deceleration is handled automatically with the human available to take over instantly when required. Level 3 is all safety critical functions are handled automatically. A human's available to instantly take over when required and that's when it gets fuzzy. And 4 is, as you mentioned, within a specific designed domain, whether that's a village, town, or a known course and known conditions, it can autonomously drive itself without human interaction. And then Level 5 is, as you also said, was anywhere, anytime that you like. And hopefully, if you tell it to navigate across the ocean, it doesn't try to do that. That kind of thing. So, the thing is about those levels, I guess, from a usefulness perspective, things get blurry for me between three and four, because I think about cutting between autonomous- Like autonomous acceleration, deceleration and most safety critical functions. And then, oh, but you may need to jump in at any moment. And the problem is that human reaction time, if you're paying attention, is already atrocious. If you're not paying attention, your reaction time is going to be even worse. It's like relying on a human to take action in a safety critical situation. In the area that I work in particular in control systems engineering, if you've got safety critical functions, you've got a dedicated safety controller that handles that in sub millisecond performance and everything's fail safe. And you don't trust a human, you know, don't trust a human to do a machine's job, right? Matrix. So, the thing is that that's where it gets a bit blurry for me. It's like level three is not a level that's very useful if it's useful at all. So, that's a bit odd. So, I think when technically people talk about autonomy, they actually really mean it starts at level four as a minimum. But level five is really what we think about as true autonomy. It's like, take me up to the mountains and then just, you know, cross your legs and sort of like fold your arms, kick back and let the car take you kind of thing. Because the other thing about level four that I find a bit weird is the definition of a design domain is also fuzzy. And I understand when they wrote this, they were trying, they were intentionally being generic, but technologically speaking, it's not terribly useful because, you know, there's all sorts of ways you can bend that to say, I have level four in quotes, autonomous driving when, you know, it all depends on how you define your domain that you're that's going to be driving in. So anyway. Yeah, I think that's I think that's a critical point. And I think where I think most people who are talking about autonomy are talking past each other is the quote you had from Intel was for level five. So that would be everything soup to nuts. The engineers need to sign off on this can do anything. rain, hail, sleet, snow, anywhere in the world, everywhere, all the time, no human. That is wildly different than saying San Francisco during the day with no rain on these streets we've already mapped. And functionally to someone inside the vehicle, someone who took a level five car from someone else and just plopped it in San Francisco could perform just as well as that level four car I just described. But you take that level four car a few miles down the road to Oakland or to Redwood City and it wouldn't know what to do. Now if you live in San Francisco, you don't care. You get to and you just need to get from one place to another. A level four car is autonomous to you. And that's where we're going to see all these companies threading the needle is figuring out what is the minimum amount for level four to make a customer, to provide value to a customer in a meaningful way. And what you're seeing are folks like Waymo and others as an engineering practice to just chop off entire classes of the level five problem. What if it's snowing? Nope, the car won't run when it snows. What if there's a hailstorm? Nope, we're not dealing with that now. What if it's snowing and fogging? Nope, not going to deal with that now. So eventually those problems will have to be solved. And that's the problem with these predictions. Because no one's even working on those problems really, in earnest, I'm sure some people are, but in earnest, those are not even edge cases. They're just so extreme that people are not caring about them that I don't think we're really going to get to level five for many, many-- potentially a decade or plus, maybe two decades, where you really could drive anywhere. I think that actually is not going to happen. I think what will actually happen is the cars will, in those cases, pull over and not drive. And you will see in certain remote parts of the world, no autonomy happening for decades. Now that doesn't mean we won't have self-driving cars in San Francisco in the next year. It just means that they may be on certain streets and if there's too much traffic, they won't dispatch. They'll dispatch at certain points in the day and when it's not raining. And so it comes down to, in my mind, figuring out what are the set of things that the cars can do under certain conditions. And Tesla may choose to make that happen on highways first. So for the long distance drive, I did some calculations and over 99.9% of the drive would be highway for the West Coast to East Coast trip. So while that sounds impressive to go West Coast to East Coast, it's really lots of miles of highway. So if you have a really good highway system your autonomous package, then you'll be majority covered. And then you have to figure out how to exit the highway for the superchargers. So you need to do a very little bit of mapping potentially for that. And hopefully you're going to be taking the easiest exits. And then when you do get into the cities, you're going to take exits to sort of avoid the deep city. And so I think it's very possible they will hit that goal of showing that demo. I don't think that software will roll out to customers at that point, but they will achieve a car driving without someone interacting with it. But that's almost kind of pointless. It shows that there's increased autonomy on the highway, but it's not... It will be level four under the constraints that this describes, but it won't necessarily be useful to people right away. And so I think you're going to see it coming from both ends. doing it in specific cities like Waymo and Cruise and others are doing. And then you've got Tesla and the other automakers who are approaching it from the complete package and they will be slowly integrating more and more functionality to regular driving. And the potential risk is that people, as you said, will atrophy and not be able to understand when their cars are able to do this. And I think the interfaces and the ability to turn on these functionalities will need to be really improved to the point where people can't get themselves in a bad situation without really, really overriding the system. Yeah, and I actually also think that we're going to need some better granularity in the definition, and it's going to need to happen in the next few years between level 4 and 5, and it's going to... maybe there's a better stand out there that I'm just not aware of, quite possible. But we need to be able to say, yes, well, we can do level four, we can do level five, or we can do level four within this subset of environments. And that needs to be standardized because people are just going to get too confused and it'll be just, people will not be able to understand the differences of different vehicles and what autopilot versus autonomous. That's another thing, actually, I really like the name autopilot, because it was intelligently chosen, because autopilot systems on aircraft, for example, is the best analogy, which is, I think, where they ripped it off from. Yeah, it just keeps it straight and level. That's it. And that's what autopilot does. And it's not really autonomous driving. It was never meant to be. So I'm excited for the future potential. Absolutely. But honestly, as it stands today, it has so far to go. And I think my biggest concern really is around how much processing power it's actually going to take to reach that, how much neural network training it's going to take. But the high level functions is really the part where you have rules and a rule set to follow as an autonomous driving. But you have to be able to bend the rules and how you actually code a machine to handle that human sort of depth of the scenario of like, for example, I'm going to move across the solid white line in the center of the highway to make more room for a cyclist because there's a big truck coming straight up my a**. So if I slow down, that's too dangerous for me. If I don't move away from the cyclist, then that presents a risk to the cyclist. There's no oncoming traffic. The road ahead is clear, so it's safe to do that. Codifying that is incredibly difficult. There's a whole set of rules and it doesn't solve the other problems regarding like philosophical issues, which is even more problematic. I know some people pan the movie, but I actually kind of liked iRobot. And when the S4 jumped into the lake after Will Smith's character and another vehicle were sinking and they were about to drown, the S4 assessed that he had a better percentage chance of survival than a young girl in an adjacent vehicle. So he- So the S4 rescued him rather than her. And the movie sort of concluded with with Sunny, who was a modified S5 robot, to have a different outcome because he was in- Yeah, he was- The robot was instructed to save the girl, despite the fact that his odds of success were- Would have been better the other way around. It chose a different outcome. And that's the sort of problem that's fascinating, and it's going to be incredibly difficult to solve. And I'm not entirely sure it can be solved because the depth of reasoning has to be so vast. I don't know. I hate to say it's impossible because I suppose anything is possible, but with artificial intelligence, it's quite possible, but maybe at some point in the future. But it just it feels like it is just so far away. Yeah, I think that I think from speaking to some people who work in this field, the and practically working on building these systems, they're not really working on the ethical portions yet because the systems are not capable enough of even doing that. They require the human to take over or they're expecting the human to take over in points where they reach sort of a threshold of safety that the system's uncomfortable with. So those decisions have not yet really been coded into the systems. So I think we'll find out about that much much later when we've got accidents happening and we'll have to debug what went wrong and as a culture and a society in different places have rulemaking around that and I think also the makers of these vehicles and systems will be asking for more guidance and I know certain countries are starting to provide some guidance around how to think about loss of property or damage to human life and who do you protect but I think that the the earlier point about the challenges of when you break the rules and all, some of the challenges that human intuition has, again, those are, it is possible to safely sort of just stay far enough away from the cyclist that, you know, the person behind you is legitimately in most, I believe, at least in the United States, they would be required to slow down for you because they're behind you. And so from what I've observed and what I've seen, many of these cars drive in such a a safe and slow way that they actually don't really behave human-like at all. They actually are so respectful of keeping boundaries around people, never running yellow lights, always coming to complete stops, that they actually cause problems for human drivers that are interacting with them. This is what Google and Waymo have been spending a lot of time on, is that they now know how to drive and follow the rule book, but they're learning how to behave more human-like. That is a very long process, and if you're doing that, as you said, in a rule-based way, you've got a lot of rules to build and a lot to validate. The promise of the deep neural networks and the AI-type learning is that you wouldn't actually codify those. The problem is not many of those systems are actually in use in any of these higher-level functions. They're currently mostly used for lane detection, object detection, free space detection, much more the perception problem. So I think most people actually cannot make a good estimation of when this will really happen, except to say that to get to a level like a Waymo is possible within a couple years if you're sort of starting fresh, and also that the necessary requirements but not sufficient of understanding where you are in the world, being able to stay in the lane, being able to accelerate, being able to take exits, those are all important prerequisites to being able to even worry about these higher level decision-making components. And so that's what they're focusing on now. I just honestly don't think that many people yet are working on sort of the trolley problem logic of many of these cars. And unfortunately, or fortunately, those are not actually going to become important decisions yet, because even if you're in an autonomous taxi in one of these cities in the next couple years, if it freaks out it's just going to pull the side of the road and put its blinkers on and not make an extreme decision. And I think that's going to be interesting to see what happens when all these cars are pulling over to the side of the road every day because they don't know what they're doing and whether or not cities find that to be not good enough for level four or good enough for their streets. But the same way like automatic emergency braking, many cars have it. I don't think many people know what the constraints are. Does it always stop? Does it stop for a deer? Does it stop for a buffalo? Does it stop for a person? Would it stop for a truck? You really don't get to test many of those systems unless something goes wrong. The interesting thing for the autonomous type systems is people will be using them in a much more active way because they'll be experiencing it as it's driving them. But having actually driven autopilot, I think people have a very strong preservation of life desire. And even if you are letting the car drive, as soon as you feel it sort of behave a little weird, you put your hands back on the wheel and you don't let it just drift away anymore. You pay attention. And I think for many of the people who have autopilot, you speak to them, they, they know the constraints of it very well because they use it and it is a life or death matter. Um, when you first try it out, I think people are a little bit more reluctant and they, uh, might just sort of be blase and take their hands off the wheel. But as soon as that car kind of darts across the lane a little bit, their hands go back on the wheel. And I think you would see. Owners of autopilot crying foul or saying I never use it. It's too unsafe. But what you actually find is many of the people who have autopilot love it and use it all the time. And they've learned what the limitations are. Just like you learn what the limitations of the traction of your vehicle are. If you've got low profile crappy tires, you're not going to be taking turns at 60 miles an hour. And if you do, it's very unsafe and you might get into an accident. So I think many of the same. Characteristics that have led people to understand what their cars can do will happen with the semi-autonomous features. I just hope that people, um, don't believe they're fully self-driving to the point where they get into an accident, they're in the back of their car and they can't take over. That I think is on the manufacturers to make sure that people are physically in their seat, looking forward, ready to take over. And that's probably the best they can do until they really can let the person get out of the seat. But I think getting out of the seat is at least a decade away. - Yeah, I absolutely agree. And I think that there's a lot of hype and a lot of excitement about it. And again, excited for the future potential of it. but it is certainly further off than, than I don't believe the hype maybe is the, is the, the, the caution on that, but all right. Okay. So we should probably wrap at this point. I, I want to just quickly talk about some of Tesla's future challenges and, and leave it there. So I guess from, from just looking at where they're at, I think they need to keep pushing battery technology forward and to look to the next thing beyond lithium, I think, or either that, or they have to either reduce their reliance on lithium or they need to secure access to it a bit more, I think. I think charging rates still not competitive or comparative is perhaps a better way of saying it with gasoline. It's still a big hurdle for a lot of people waiting for 30 minutes or more to have their car charged is a very different. They have to change the way that they manage their travel and so on. And for some people, they're prepared to wear that because it's not a big inconvenience. but for other people they're saying, well, that's just ridiculous. I can fill my car up in five minutes or less. Why would I with gasoline? Why would I want to wait half an hour to an hour to fully charge my Tesla? So that is something that needs to be pushed forward. The battery exchanging prototype that they did, the Tesla tried didn't really work out and they kind of shut it at very very quietly. Sorry. And but that's a shame. Maybe maybe that was a dead end too difficult from engineering point of view perhaps, but in any case that the charging rate problem still needs to be addressed. I don't know whether they look at buffering capacitors, changing battery geometries. There's ideas of like, you know, combined, like actually having a pumped electrolytes that's pre-charged. There's a whole bunch of different options, but whatever the solution is, they need to be working on that. I'm sure that they are. Again, maybe different battery technology would help with that, maybe. Autopilot, obviously there's a lot of work to still be done on that. I think neural nets will only take it so far. Algorithmically processing power to get that last 20%, as you mentioned, it's, it's going to take a lot longer, but that's okay. Um, the, I don't think everything that I've seen about Tesla suggests that they're not going to give up on that. So that's, uh, they'll be chasing that harder than anyone else in the market. I think, um, the costs need to come down. Uh, obviously the model three is their first stab at, at, at that. And honestly, um, if it takes off, I can see that they may in the longer term, five to 10 years, prices will come down some more, but probably not too much more. That's probably close enough for most people. I guess we'll wait and see when they're shipping a hundred thousand cars a year or so. I guess I'll believe whatever they can produce at that point as well. And yeah, more gigafactories, more parallel production, really. I don't know, what other challenges do you see or if you have anything to add to that? Yeah, I mean, I think the ordering for me would be production capability first, uh, as the most sort of acute, um, in the longterm, of course they have to have that solved, but in the short term, that's a big risk for the, for the company. So I think that's gotta be priority number one right now. I think secondly, after that has to be autonomous and autopilot type features. I feel like, um, if you, if you don't make enough progress on that quickly, Um, They have the potential to really miss out, um, in a huge way. Uh, so I think that Tesla will have to be one of the companies that owns and controls their own autopilot and autonomy features, and hopefully will be one of the best. Uh, I think certainly. Understand what's what's necessary through some of the work they've done with SpaceX. And obviously it may not be obvious actually, but I think landing a rocket is simpler than riding a car in New York city, but it's at least people thought that was impossible and they've done that and gotten pretty good at it. So I think certainly have some understanding of what autonomy means and how to control very large expensive one-time systems. So I think that's important. I think the massive increase in production automation sort of dovetails with that idea that Tesla wants to be able to produce multiple, you know, multiple cars at the same rate that other car companies make one car. And I Especially I think eventually they want to get to ten times as many cars per square foot as other car makers which would massively reduce their costs So and I think to your point, I don't think they'll necessarily reduce the cost of the customer And so I believe they will continue to be one of the highest profit margin Auto companies which would make them one of the most valuable companies because if you apply Apple margins to car prices You have a really big company Where traditional margins for car companies are more in the low single digits Tesla gross margins are in the high 20s So if they can keep that going forward, they'll be one of the most valuable companies that could surpass Apple So I think that's sort of another big one. And then I I think I slightly disagree on the charging rate I think the way that actually gets solved is through 400 500 mile packs Because then you would be able to insert the juice a lot faster at the lower rates at the lower charge rates and Be able to give them useful range in 10 or 15 20 minutes by pumping in maybe 150 kilowatts 200 kilowatts into a much bigger pack it would be able to spread that out across the cells instead of Right now where it's sort of the smaller packs have to charge at a much slower rate So I think you I think that the goal of faster charging is important But I think they may just be able to get there with bigger and bigger packs and faster charging through those packs versus some technology breakthrough or change And yeah, I think those are those are main things obviously they need to keep adding more superchargers in stores But that's sort of block and tackling stuff. I don't think that's Really big strategic decisions to be made there. I think they just need to keep executing on that and keeping pace Okay, fair enough about the batteries and the charging rate I suppose if the price of your batteries does go down to a point where you can continue to make them the larger battery packs then that's true that is that is another way of handling it I suppose so long as the batteries remain such a significant cost to the vehicle and to maintain your profit margin that that that solution will be a longer time coming I suppose is yeah I suppose is the way I'd see that but it's going to be an interesting space to watch that's for sure and And once they open up more than one gigafactory, maybe that will become an option. So, and that could, well, you said that's gonna be announced later this year. You'd heard, is that right? - Yep, yep, that's what they've said. They plan to announce at least one or two more. - Fantastic. - Probably China and Western Europe or Eastern US. I think probably Tennessee or South Carolina or something and probably China would be my bet. - Yep, okay, fair enough. Cool, so just in closing about Tesla, I think that they have changed the world already and they've changed this space significantly. And I think about Tesla, I think had a part based on my understanding of Nissan or Nissan, Nissan, I can never pronounce it correctly. Anyway, then for the Leaf, the Nissan Leaf. Yep. And the Volt. And the Volt. Yes, that's right. And subsequently the Bolt. they were either in part or debatably mainly due caused those programs to move ahead or to even be created. And Tesla have popularized electric vehicles. And if you look across all the vehicle manufacturers in the world, they're either currently working on or they have electric vehicle programs in progress. And those that don't have models, electric vehicles in the market, in the field yet, I should say, will have probably in the next five years. And that is quite amazing, actually, when you think about it. And and that just brings it back to Tesla's whole motivation, which was to accelerate the transition to sustainable transportation. And they may only be making a few hundred thousand cars a year, but the impact that they're having goes far beyond the cars that they make. And I think that that is the most amazing thing, to be honest, about Tesla so far. Yeah, I agree. I feel like the amount of impact they have relative to their actual production Is pretty immense and I think it's pretty hard to deny that Tesla has woken up other car makers to take seriously Ev's and and make them possible sooner than they would have otherwise I don't think there was much evidence that other car makers were really pushing forward on electric vehicles before Tesla really started gaining much attention with the Model S and Eventually Model X and I think the Model 3 clearly woke a lot of people up that there are people who want Electric vehicles that are not crazy green eco people and That people don't want gasoline cars they just need a good car at a good price that fits their style and they don't really care that it's gasoline and I think as soon as people really understand that Almost everything about an electric car is better except for some of the charging things But only in the charging states when you're doing long-distance, but to wake up every morning and have your car fully fully available with all of its range if assuming you have a plug at your home is pretty powerful and To then imagine that the car will take you to work It's pretty great. And then eventually you won't even have to own the car. I think is it's very exciting that for 50 years or so not much has really changed with cars obviously there have been incremental improvements but nothing fundamental has changed and Tesla is the representation of that change that's happening and it's just really exciting to watch. Awesome, alrighty, well if you'd like to talk more about this you can reach me on Mastodon at [email protected] or you can follow @engineered_network on Twitter to see announcements about all the shows on the Engineer Network which you can find at engineer.network. Causality recently has taken off. It's a solid podcast I did that looks at the cause and effect of major events and disasters in history. So if you're a fan of Pragmatic, you may like it as well. So be sure to check that out. If you'd like to get in touch with Caleb, what's the best way for them to get in touch with you, mate? Yeah, probably on Twitter @theteslashow or theteslashow.com. Awesome. And if you'd like to send any feedback about the show or the network, please use the feedback form on the website. That's where you'll also find show notes for this episode. If you're enjoying Pragmatic and you want to support the show, you can. Like some of our backers, Ivan and Chris Stone, they and many others are patrons of the show via Patreon, and you can find it at patreon.com/johnchidjee, all one word. So if you'd like to contribute something, anything at all, it's all very much appreciated. A special thank you to all of our patrons, a big thank you to everyone else for listening, and as always, so thank you very much for joining me to talk about Tesla. I have thoroughly enjoyed it. I don't often get a chance to talk to people about Tesla. I've been a bit of a fan of your podcast for a while, so I really do appreciate you coming on the show and making the time, Kayla. Thank you. Thanks, John. Always appreciate it, and it's a lot of fun. Cool, and I hope you get your Model 3 soon. Me too. I'll be posting lots of photos. I think you'll find out about about it on the podcast for sure. - Fantastic. (upbeat electronic music) (upbeat music) [Music] (upbeat music) [MUSIC] (upbeat music) (upbeat music) (upbeat music) (upbeat music) (upbeat music) [Music] (explosion)
Duration 1 hour, 25 minutes and 21 seconds Direct Download

Show Notes

Battery:

AutoPilot:


Episode Gold Producer: 'r'.
Episode Silver Producers: Chris Stone and Eivind Hjertnes.
Premium supporters have access to high-quality, early released episodes with a full back-catalogues of previous episodes
SUPPORT PRAGMATIC PATREON APPLE PODCASTS PAYPAL ME
STREAMING VALUE SUPPORT FOUNTAIN PODVERSE BREEZ PODFRIEND
CONTACT FEEDBACK REDDIT FEDIVERSE TWITTER FACEBOOK
LISTEN RSS PODFRIEND APPLE PODCASTS SPOTIFY GOOGLE PODCASTS INSTAGRAM STITCHER IHEART RADIO TUNEIN RADIO CASTBOX FM OVERCAST POCKETCASTS CASTRO GAANA JIOSAAVN AMAZON

People


Caleb Elston

Caleb Elston

Caleb is a passionate Tesla follower and has run a wonderful and informative Tesla-centric podcast, has been on several Fremont factory tours and attended some of the recent Tesla events.

John Chidgey

John Chidgey

John is an Electrical, Instrumentation and Control Systems Engineer, software developer, podcaster, vocal actor and runs TechDistortion and the Engineered Network. John is a Chartered Professional Engineer in both Electrical Engineering and Information, Telecommunications and Electronics Engineering (ITEE) and a semi-regular conference speaker.

John has produced and appeared on many podcasts including Pragmatic and Causality and is available for hire for Vocal Acting or advertising. He has experience and interest in HMI Design, Alarm Management, Cyber-security and Root Cause Analysis.

Described as the David Attenborough of disasters, and a Dreamy Narrator with Great Pipes by the Podfather Adam Curry.

You can find him on the Fediverse and on Twitter.