Professor Markus Lienkamp, of Technische Universität München (roughly the MIT of Germany) was the only academic who presented at the Automotiv Elektronik Kongress in Ludwigsburg. His presentation was The Status of Electromobility in 2017, or How Tesla Will Not Win . As he put it: Americans always want to conquer the world, but German OEMs have a chance to present products as good. German Electric Car Sales The above table shows the driving range for each vehicle (in blue) and the number of vehicles sold (in green). It is interesting that even in Germany there are only a few vehicles sold per quarter and so both bars on the graph can use the same scale. Tesla has the longest range and eyeballing the graph sells as many electric cars as everyone else added together. So far Tesla is winning, but that is a bit like judging a marathon by who is ahead after the first quarter mile. So let's look at Markus's reasons for predicting that Tesla will not win and German OEMs will. And to wrap up, I'll look at what I consider a major omission from the analysis. Bodywork The Tesla has aluminum bodywork. But that is not a Tesla invention, Audi has been using aluminum bodywork for several vehicles and has manufactured 250,000 of them. Major manufacturers such as BMW have a lot of competence in mixed-material body structures, which means using carbon fiber, steel, and aluminum as appropriate. When I lived in France I had a Renault Espace, which had fiberglass bodywork on a steel frame. One reason that steel is so great for bodywork is that it takes a lot of energy to bend steel, compared to aluminum or fiberglass, meaning that a a large amount of energy in a crash is absorbed by the need to crush a lot of metal before you get to the passenger compartment. One advantage Tesla has had up to now is that they only build electric vehicles. So they have not had to take the body of, say, a VW Golf and add batteries and an electric power train to the same chassis that was originally designed for ICE (that's internal combustion engine). Tesla can thus have underbody batteries to keep weight low and central. In future designs, all manufacturers will design underbody batteries as standard and so this won't be a Tesla advantage any more. The Electric Machine of the OEMs There are three technologies for electric propulsion: Induction Machine (IM) Used by Toyota and Tesla Permanent Synchronous Machine (PSM) Used bt BMW, BYD and Nissan Externally excited synchronous motor (FSM), I guess that comes from the French since it is... ...used by Renault The PSM is the best but it is more expensive. I don't know if the red for manufacturing is different from the red for price, or whether this is the same thing twice (it is more difficult to manufacture so it costs more). Anyway, apparently the IM has no specific Tesla advantage. Without knowing more, such as the cost difference, it is hard to tell. If you put "jet turbine" on the chart it would have a lot of power, but have some pretty major disadvantages on a car, despite working so well on passenger aircraft, but if you picked the right features you might get it to only have a couple of red boxes. When Markus moved onto the inverters, again there are pros and cons in the choice Tesla has made. It has a larger cooling area so better overload potential, and it has a lower price. But it has 14 IGBT (integrated gate bipolar transistors) in parallel and large size so...survey says: No specific advantage for Tesla. And anyway, in the future, everyone will buy these power modules from the same semiconductor manufacturers and they will switch to (probably) SiC (silicon carbide). Batteries There are several cell properties you want in a good battery, and as usual, they work against each other as in the classic tradeoff ("fast, cheap, good: pick any two"). These are: Safety Efficiency Weight Cost For example, large cells (known as pouch) waste less space and so have better power density. On the other hand, they have less inherent thermal dissipation and so a higher risk of thermal runaway compared to small cells (cylindrical, which is what Tesla uses and what their gigafactory in Nevada produces). The TUM have built a battery model to look at the impact of design decisions on battery performance. It even takes into account things like the amount of scrap during manufacturing (small cells are better) and the fact that with small cells, you can save the neighboring cells if one small one runs away and has to be shut down. Looking at how costs change over time, it looks like: Tesla went with small cells, which was a good decision at the time. But in 2018, Renault and others will use large cells, which are cheaper, and only large cells will ever be cheap. So, of course, the professor has decided that the Tesla choice is wrong going forward: When you take small cells (4.5Ah) end up around $130/KWh. Will go down to $100 per KWh. But if we take large cells can get down to $85/KWh. So $4K per vehicle. A hybridized 84V onboard network another $7-8K. All in all, recommend to build large cells if safety and efficiency and production quality can be maintained. Tesla has cost disadvantage of about $2K compared to large cells. We think that this is a big number in automotive if you can’t provide an customer benefit. Software Development Tesla has decoupled software and hardware development, and has over-the-air updates. If you look at the above chart, you can see that Tesla has about 3-6 months between software versions (on the right of the timeline down the middle) and Mercedes needs one to three years (on the left). Updates are pretty much a requirement for autonomous driving, and it is a big advantage for Tesla to have every vehicle equipped with the latest software generating valuable real-time data. Here's one example. In 2015, researchers hacked into a Jeep (read my post Automotive Security: a Hacker's Eye View ) and the response was that 1.4M Chrysler owners had to contact the company for a USB stick with the update on it. In 2016, researchers hacked into a Tesla model S. Within just 10 days, Tesla deployed an over-the-air software update that addressed the security issues. Conclusion Well the conclusion is in the title. Tesla will not win. Tesla has no advantage in conventional automotive manufacturing. But it has more experience in over-the-air and ADAS data collection. They may have a $2K disadvantage in battery pack cost due to technology. I think that the case is very unproven. Things I think are not sufficiently taken into account: Alll the non-Tesla companies have expensive stranded assets and big plants for building engines. They can't even push electric cars too hard since they need people to keep buying ICE cars for the next several years. This feels like an Innovator's Dilemma type of transition, such steel minimills coming along and the big steel companies saying "they'll never scale to blast furnaces as big as ours, and rolling mills for high-quality steel like we have." Tesla is not just building batteries for automotive but also for solar power. The cost of Tesla's batteries comes down every time people put in a solar installation using a PowerWall. Plus, for now, even disregarding that, their volumes are a lot higher than anyone else's. I have no specialized knowledge about Tesla's battery costs versus other technologies. They may well be slightly higher. But they may well need smaller packs due to better software. In fact, better software is probably the most important thing. Lord Rutherford had a saying that "all science is either physics or stamp-collecting." Well, autonomous electric vehicles are either all about software or stamp-collecting. Even Tesla's extended battery life comes from better software. Tesla came out of nowhere to show you could start a car company from a standing start, provided the key capability was software not manufacturing prowess. I'm quite prepared to believe some of the statements by the big car manufacturers that scaling to millions of vehicles a year is not easy. But then if you were the CEO of General Motors, that's what you would say even if you were terrified it wasn't true. Besides, for now, Tesla even has the advantage of volume—volume of high-level ADAS all-electric vehicle manufacturing. I'm also quite prepared to believe that no major automotive company will get their software capabilities high enough from a standing start. My conclusion: German professors (and car companies) hope that Tesla will not win. Tesla Model 3 Since this was presented, last week Tesla launched the Model 3. This is the first Tesla car priced for the mainstream. It is hard to underestimate the extent to which Tesla's cars are perceived as sexy, and other electric vehicles are just functional. It is a bit like the iPhone versus some no-name brand. The specs may not even be that different. But iPhone and Tesla are aspirational brands (to be fair, so are BMW and Mercedes, at least for ICE vehicles). Tesla might not "win" in the sense of having the biggest market share, in the same way as iPhone doesn't win (Samsung has double the market share). Tesla's financials are not good at current volumes, and I've not looked into it enough to see if that is likely to improve with scale. Whether they make money or not is probably not the biggest thing—they have already completely changed the perception of what is possible with an electric vehicle and how close we are to autonomous driving in more and more circumstances. Paper There is a full paper/ebook with the same title . Or click on the cover to the right. Sign up for the weekly Breakfast Bytes email:![]()
![]()
↧