Search

NHTSA Investigates Autopilot, Pony.ai No Go on IPO, Ola Gamechanger - SAI Newsletter 32



We will go short and sweet with this week’s newsletter. My family just got back from Watertown - a resort area just north of downtown BJ. With restrictions popping up all over China we decided that it best to keep it local and play it safe so we scrapped our trip to Yunnan Province. The resort was great and last night we were able to watch the sunset on the Great Wall. I’ve said this before but it bears repeating, the Great Wall is one of the COOLEST things most people will see in their lives. For those in China (and outside frankly), I recommend coming to check it out. Some parts of the Wall are more popular than others so I’d also recommend that you try to NOT go to the most touristy of them all and avoid going during a national holiday.

A pic of the sunset from Tuesday evening on the Wall. I also just returned today from the Baidu World AI Conference and will talk about that a bit more in a link below. For those who follow the mobility space closely, two of the big items we’re following is the NHTSA investigation into Tesla vehicles hitting parked emergency vehicles in the US. I also talk about it a bit more down below. The other item that’s been moving markets is the NIO crash from a few weeks back. More details are starting to trickle out about the accident that killed a young Chinese entrepreneur. NIO’s ADAS (autonomous driving system) is called Navigation on Pilot or NOP for short. The driver seemed to overestimate the system’s capabilities and was lulled into a false sense of security. NIO clearly claims that their system is L2 so there shouldn’t be any liability on their side unless the investigation finds that there is a flaw in it. Still nothing definitive about the cause although now there are firm accusations from the victim’s estate and firm denials by NIO. Those events have played a large part in both company’s share prices taking a bit of a nosedive over the last week or so. Lei and I will host our Twitter Spaces an hour early this week so that it won’t interfere with Tesla’s AI day. For those interested in joining, we’ll be on Thursday – 7:30pm EST. Please do join us. For those who can’t join in real-time but are interested in listening to this week’s or past episodes of China EVs & More, you can find them by scanning the below QR code: TESLA NEWS - Tesla’s Autopilot has raised enough red flags that the National Highway Transportation Safety Administration or NHTSA as they are better known, has opened an investigation into the feature. Here are the hard numbers. There have been at least 11 accidents as of 2018 that have involved Tesla vehicles with Autopilot activated crashing into parked emergency vehicles that had their flashing lights on. There have also been some civilian accidents, a few of them serious including this tragic fatality detailed in the linked NYTimes article. In this particular case, there are still no definitive conclusions as to the cause or even what really happened but what is clear is how some Tesla owners overestimate the system’s capabilities. Although Tesla recommends, like other OEMs that sell vehicles with ADAS capabilities, that the system ONLY be used on highways and not local streets there are numerous examples online of folks using Autopilot to drive around areas with busy intersections that include traffic lights and merging onto oncoming traffic. On top of this, Tesla’s Driver Monitoring System (DMS) is virtually non-existent and/or can be gamed easily. This technology is still in its infancy so erring on the side of safety, other people’s in particular, should be prioritized over all else. How that’s made practical is still a work in progress but one way is to make the DMS much more bulletproof so that it can’t be gamed or manipulated. I think that would be a good start. And if the automakers can’t self-regulate themselves, then federal entities like the NHTSA need to do more themselves. IN THE NEWS - Getting to Job #1 on the cheap. Companies like Arrival, Ree, Fisker, Canoo, and more are all hoping that either building their products using smaller factories or outsourcing completely will help them get their vehicles on the road much more affordably vs. going the traditional route of investing in a massive new factory that could cost >$2B. One thing that’s worth noting though is that these companies are also not trying to rely on the type of volume, at least initially, where a large factory would make a ton of sense. The good thing for these companies is that they can look to China and companies like NIO & Xpeng who both originally completely outsourced their manufacturing. They are both adding capacity now and in XPeng’s case, decided that it made sense to build their own factory, which it did, and is now looking at adding another. They can also look at Mercedes, Bimmer, and Toyota in the EU who’ve successfully outsourced some of their manufacturing to Magna. I am a bit skeptical that all of these companies will succeed with these models. It creates a ton of flexibility but at the end of the day, the ultimate decision they need to make is whether or not building should be one of their key differentiators and hence difference makers. BTW, Job #1 refers to the first production vehicle of a model coming off of the line. - Baidu introduces an L5 robotaxi, an AI chip, and some consumer products? I wasn’t sure what to expect since this was my Baidu World Conference. What I got was an invitation to a movie theatre not far from campus due to COVID restrictions. Baidu came out strong with the reveal of their L5 taxi. I wouldn’t say it’s a good-looking vehicle and passengers will need to get used to not being able to see out of the front window ….since there isn’t one! They finished strong with the introduction of an AI chip, something desperately needed if they plan to shake their dependency on non-Chinese tech. Overall, it seemed like they wanted to re-brand themselves as an AI company that utilizes its skills and expertise to help industries across the board. Whether it’s agriculture, water filtration, autonomous vehicles, or consumers products, through AI Baidu plays a large role. If I am being honest, there were some pretty awkward moments and co-hosting the event with CCTV was a bit strange. Also, it was surreal to sit in a theatre watching as I would’ve loved to see, touch, and hear the L5 taxi in person. Alas, this is the world we currently live in. TRENDING ON SOCIAL MEDIA - Pony.ai comes to a screeching halt. Pony’s plan for a US IPO later this year was shelved after they couldn’t coax assurances out of the Chinese govt that it wouldn’t have any issues post-IPO alá Didi. There are a number of investors & the SPAC owner VectoIQ that probably need a hug right about now. Remember that the Steve Girsky led VectoIQ brought GM the Nikola deal after fully endorsing the company. They’re still trying to wipe the stink off of themselves for that dog. The Pony transaction would’ve likely helped people forget about Nikola but that’s not happening now. Pony was likely going to be THE premiere AV/robotaxi startup coming out of China and the rumor was they were looking at a $12B valuation. From a positioning standpoint and if you asked them, they’d say they’re an American company. Kinda like TuSimple is an American company. Wink. Wink. This likely means a listing in HK or Shanghai at a lower valuation for Pony and the early investors won’t likely get paid out for > 3-4 years as they try to raise additional funding to stay competitive. Unless they dump their ownership stakes out on the secondary market that is. With the US capital markets off-limits to many Chinese companies now from both the US & China sides (more on that in the Get Smarter section), it will be a battle of attrition for many of these companies since not all of them will get that continued funding. And if there’s one thing we know, getting to L4-5 will be VERY expensive, if at all possible. - Deep dive into Waymo’s challenges that spotlight the challenges of the sector writ large. Looks like manufacturing is indeed a bottleneck for Waymo as well. Advantage: OEMs. That’s not likely sustainable unless the HW/SW stack is better at these OEMs and they’re able to consistently invest in putting more data vacuums, I mean robotaxi pilots on the road. One final observation, this article seemed to focus on the US AV sector, but it should be noted that there are some viable challengers to Waymo here in China. We highlighted one of them in a post above. - Li Auto exploring building abroad. As I’ve said before, Li Auto hasn’t followed its contemporaries when it comes to company strategy so why should it start now? There’s a rumor that they’re looking to build a factory outside of China in order to increase the chances of successful international expansion. This makes a TON of sense. First, none of the Chinese automakers will succeed long-term or even get close to generating significant sales if they only plan to ship product from China. The logistics and inventory costs, along with the tariff would limit potential straight away. By building in the EU it avoids the tariff and logistics costs and can price their vehicles aggressively to gain share. This also means that shipping vehicles from the EU save them ~25% in tariff costs since the EU -> US tariff rate is 2.5% vs. 27.5% for China -> the US (thanks to the Trump administration). The EU & US markets are large enough to where building capacity would be economical and provide some goodwill since they’d be hiring and building locally. It would be similar to what BMW, Mercedes, Hyundai, Toyota, Honda, and others who all now build locally in the US. With BMW, they build more cars in the US than they do in Germany. That’s not to say that Li’s competitors aren’t also running the numbers for building outside of China. They have to be. But who will actually work with a local municipal govt in the EU, Canada, Mexico, or the US to try to get to numbers that make sense for them remains to be seen. What we do know is that building locally is a key component to Tesla’s strategy and look at their sales growth over the last couple of years. GET SMARTER - The SEC wants to effectively close the door on Chinese companies that list in the US via a Variable Interest

Entity (VIE) structure. There was always a problem with these VIEs since the Chinese govt doesn’t officially recognize them but simultaneously will not let foreign entities own Chinese domestic companies. Despite that risk, the VIE worked swimmingly well for companies like Alibaba, Baidu, NIO, Netease, Pinduoduo, and many other Chinese companies. That’s until the Didi IPO. Now the tougher stance from both the Chinese govt. and the SEC has made it much more difficult for a number of companies that collect Chinese citizens’ data to cash out. Make that all but impossible, at least for the rest of this year. I don’t see that changing anytime in the near future either so what we’ll see is more Chinese companies listing on HK’s Hang Seng and SH’s StarBoard. This will also make it more difficult for those companies to do business in the US and likely lower their valuations. JUST THE NUMBERS - Between $15M and $154M. That’s how much Neolix, who we mentioned in last week’s newsletter as expanding their use cases in order to create more opportunities to generate revenue and bump of the utilization of their products, was able to raise from lead investor Softbank and a few others. Let’s just say I am thinking it’s not $154M… INTRODUCING - The Ola S1 & S1 Pro. Many my readers get from point A to B using vehicles with 4 wheels but a large % of readers also predominantly use 2 wheels to get around. In particular, Southeast Asia and South Asia are moped strongholds due to a combination of the moped’s affordability and a lack of infrastructure for passenger vehicles. That’s why the EV revolution in those regions will begin and end with electric mopeds and Ola Electric wants to lead that revolution. For those that aren’t familiar with Ola, originally they were a cab/ride-hailing company started in 2010 in India. In 2019, they spun off their electric moped division into Ola Electric and acquired a Dutch startup named Etergo in May 2020. The combination of Ola’s and Etergo’s work was introduced last week as the S1 & S1 Pro. With all the goodies you’d expect from an electric moped including a 7” digital dash, its own in-house OS, and under the seat storage, the real pleasant surprise is the price – the S1 will start at a mind-boggling low price of $1,350! If the quality and reliability are there, these things should sell by the 10’s of thousands. Globally. —— This weekly newsletter is a collection of articles we feel best reflect the happenings of the week or important trends that have effects on the automotive and mobility sectors here and in the US, we also provide a point of view that we hope educates and sparks debate.

Sino Auto Insights is a Beijing, China-based market research and advisory firm that specializes in assisting companies analyze, strategize, and develop products and services that will shape the future of mobility and transportation. Members of our team have experience working in Detroit, Silicon Valley as well as here in China across multiple sectors and functions as entrepreneurs as well as working at larger companies like Apple, Google, Amazon, GM and FCA, and many others.

44 views0 comments