SAI Weekly 27 - 24: WSJ Shines a Light on FSD, Stellantis Looking at ALL OPTIONs to get Healthy, BYD + Uber & BYD + Canada
It’s August. The summer seems to have flown by. Crazy thing is – this summer in MI, although hot, hasn’t hit crazy triple digits that often so it’s been, dare I say, bearable?
Still working the kinks out of this reset to Wednesday and as you can see, it’s working out quite well – NOT.
That said, I’ve been pretty slammed on helping out with some local startups here in the metro Detroit area which has taken much of my time this week and likely the next few.
I was interviewed last night by CNBC to discuss Toyota’s earnings as they came hot off the press, you can check out the interview below. In general, if you look at Stellantis, VW Group, Tesla, Ford and now Toyota’s recent earnings announcements – it seems the good news for most OEMs seem to be in the rearview mirror. (video courtesy of CNBC)
What I do like is how Tavares looks to have a renewed confidence in the Stellantis strategy, no doubt partially boosted by their partnership in LeapMotor. Volkswagen is also starting to embrace the reality of their situations in China, Europe and the US.
But the VW Group drama seems to ooze with internal politics, large, opinionated shareholders and constraints that don’t allow for the implementation of real drastic measures for correcting what Blume, I am certain, knows needs to happen to right the ship.
It recently it hasn’t stopped him from trying to make a few bold moves like shutting down factories and reconciling VW Group’s plan for Cariad now that Rivian and XPeng are in play. Is that enough for VW? Is having a small ‘skunkworks’ team in SoCal enough for Ford? Not so sure about either.
QUOTED / INTERVIEWED
- Had a good chat with Grant Schwab of the Detroit News about the EV demand slowdown and the challenges it poses to the D3. You can find that article here.
BIGGEST NEWS THIS WEEK
- A bit better understanding of Tesla Autopilot and Full Self Driving. I don’t actually know the difference between Tesla’s two intelligent driving systems – Autopilot which comes standard with Tesla’s and their premium version Full Self Driving aka FSD. So I let Google AI sort it out. This is what it came up with:
Combines Autopilot with navigation features to help with highway driving, including finding routes, navigating on and off ramps, and making lane changes. FSD also includes other features like Traffic and Stop Sign Control (Beta), which can identify stop signs and traffic lights and slow the vehicle down. However, FSD doesn't allow fully autonomous driving, and drivers are still responsible for keeping their hands on the wheel and being ready to take action.
Here are the basics – most automakers now offer some level of intelligent driving, level 2+. Notice I wrote ‘intelligent driving’ and not autonomous driving, but we can get into that a bit later.
They each have their own branding too. FSD, SuperCruise (GM), NOP (NIO), X-NGP (XPeng), Qiankun (Huawei), NOA (Li Auto) and so on.
You could even say adaptive cruise control, which has been offered for many years by many brands was the first intelligent driving feature, except that you needed to steer the vehicle at all times. Also, another way to say ‘intelligent driving system’ is Advanced Driving Assist System (ADAS). But what neither are is autonomous.
Here’s the autonomous definition for you:
Denoting or performed by a device capable of operating without direct human control.
But the line between intelligent driving and autonomous has been blurred. Legally and literally. Legally, with a Level 2 intelligent driving system the driver is responsible for control of the vehicle at all times. That’s why it’s inaccurate to say ‘autonomous’ because the driver still needs controls it.
And currently, no vehicle can continually drive itself without someone needing to intervene. That’s the ultimate goal for all parties involved - Level 5 – when there is no steering wheel, accelerator or brake pedal. GM’s version, the Cruise Origin was recently shelved for the foreeable future, a future that Tesla claims is much closer for them than anyone else.
Back to the basics – L1 or Level 1 is right now. You and I driving the vehicle with both our hands and feet. It’s at L3 where a driver may not need to be paying complete attention and here’s the key, it’s the level that the OEM takes on the responsibility for any accidents that occur (generally).
That’s why – with the exception of Honda and Merc who’ve both launched L3 systems to the wild – everyone else still calls their version of intelligent driving an L2 system. It’s not because they don’t believe that the vehicles can’t manage 99% of the day-to-day driving of most people, because many of them (are too overconfident) and do believe it, but they still don’t call them L3 systems because of the liability. And BTW, 99% isn’t the goal. 1.33 deaths / 100 million miles driven is where we are currently. True autonomous vehicles need track at a better rate. Much better. And when most vehicles on the road are autonomous, the number needs to trend towards ZERO.
Tesla is being accused of improperly marketing their FSD to consumers and is currently being investigated for it. It should be noted that Merc and Honda systems can’t be activated anywhere nor in every situation, quite the contrary actually, they can only be activated when certain conditions are met or only on specific roads and in Merc’s case in certain US states. With the Honda system specifically, it can be activated during traffic jams and disengages once you reach a certain speed.
There are technical differences in these systems as well. Each company has their own algorithm, control units, software and combination of sensors, cameras and, with the exception of Tesla and Jiyue, LiDAR that when combined with the other sensors, generate an image of the road ahead that’s recalculated hundreds of times / second. This ‘tech stack’ is different for everyone but does include an edge computer in the vehicle that complements what’s in the cloud.
The image is then used by the path planning system to decide what the vehicle should do. And that’s where things can go wrong. If an object is identified incorrectly, or not at all, it can lead to accidents. If we are going 10-15 MPH, the accident is often more nuisance than serious but at highway speeds mistakes can lead to serious accidents. In Tesla’s case, quite a few. Including fatalities.
I am very much oversimplifying this, but I think the above explanation is fairly accurate, if not very specific.
As to LiDAR or not to LiDAR. I believe when Tesla initially began R&D for their intelligent driving system that’s now called FSD, LiDARs spun, were large and very expensive so they made a form factor and cost decision (likely not a technical one) that LiDAR wouldn’t be used to help their vehicles ‘see.’
Fast forward to today – LiDAR are small, many of them do NOT have moving parts and can be purchased for <$500 thanks in large part due to companies like Hesai and Innovusion, two Chinese LiDAR companies that have driven LiDAR prices down substantially. Enough so that in China, you may see one on a car priced at less than $30K.
Another reason for LiDAR if you don’t believe that utilizing them makes your system better? Redundancy. This is important. And could save lives. But for Tesla, the use of LiDAR now would likely mean they’d need to almost restart their quest for L5. They’d need to redesign their future products and it would mean the data from 6 million Tesla’s on the road around the world currently would not match with any of their new products data with LiDAR.
My thoughts on this – Tesla obviously have legal disclaimers that they fully believe releases them from any liability, specifically in the case of fatal accidents. But it’s infuriating to me that NHTSA hasn’t done more to regulate and police its use because innocent folks that aren’t driving Tesla’s and hence don’t use Autopilot / FSD are still being forced to be beta testers for those Tesla drivers that do. That’s not right.
Here's the rub – I don’t know what the right answer is, but that’s also why I am not a regulator. All I know is that it seems like FSD isn’t ready for primetime, but due to the lack of transparency by both the govt and Tesla, we can’t be sure.
Let me also remind you that I used an L2 system for most of the 1.5K miles I drove from Beijing to Shenzhen back in April. Pushing forward with pilots on the robotaxi side while offering better intelligent driving systems in passenger vehicles we own / lease for ourselves is what will provide more awareness and education about what the technology can and can’t do. And I am all for that. But there’s way to go about making those two things happen without putting unsuspecting bystanders at risk.
Jumping off my soapbox now.
BTW, WSJ has done a great job with this piece of investigative journalism, for subscribers the 11 minute video is worth a watch.
- It may finally be curtains for one very infamous Chinese EV company. China Evergrande, who gained recent notoriety because property development business was accused of cooking its books, may now have another one of its businesses shutting down as well.
Two of its creditors are pushing for bankruptcy proceedings and a reorg for Evergrande Auto, the publicly traded EV business that never quite got off the ground although it had ambitions of one day competing with the likes of Tesla and other China EV Inc.
- BYD heading to Canada, eh? That’s right – the Canadian government has not decided whether it will impose tariffs, but the uncertainty hasn’t stopped BYD from making its market entry ambitions known, according to a regulatory document it filed with the Canadian govt. For those wondering, that is >80 countries and counting.
BY THE NUMBERS
- $15B: That’s how much Volkswagen Group investment spending between 2025 – 2029 has been reduced by going from $180B down to $165B. That is still a substantial sum of money but is it really about investment spending?
VW Group has announced investments in both XPeng and Rivian in hopes of having both companies bail them out of the EV mess that Cariad and their ID Series vehicles got them into the last few years.
As big as that number is – how will they change the culture? Scale is a virtue until it’s not. And right now that virtue has become a weight. The company now not only needs to get smaller because they need less staff for EV development, but they also need less staff because sales volume has drastically decreased in China.
- 17.25%: That’s how much the price of the Fang Cheng Bao Bao 5 has been reduced. It’s a B segment SUV, one from BYD’s premium brand who must be having some challenges moving the metal.
With such a drastic reduction, Fang Cheng Bao (BYD proxy) looks to be resetting the premium midsize SUV market – this increases the already enormous pressure on ABB (Audi, Bimmer, Benz) who’ve all recently said No Mas to the price war, even raising prices on their products.
This could cause some real pain for the German Three, and I think that that’s the whole point.
- 21%. Stellantis sales were down by that much in 2Q’24. High prices and high inventories and not enough incentives to spur sales leaves them in a pickle. New products still aren’t coming for some time so in the short-term, they’ll need to either reduce prices to get the inventories moved onto customers or cut production until they can bleed the inventory – knowing Tavares, it may be both.
Also, and this is important because I don’t believe it to be very easy is that Tavares has said the shuttering poor performing brands of which Stellantis has a few. How many and which ones - that’s to be determined but this shows that Tavares is leaving no stone unturned in his quest to transform Stellantis into the biggest clean energy OEM on the block. VW Group probably needs to take a long, hard look at all the brands to see if all of them are worth keeping as well.
But don’t feel too bad for Stellantis, they got while the getting was good by pricing Grand Wagoneers at well over $100K for the last few years. That eye watering pricing forced customers to look elsewhere as interest rates increased. It looks like LeapMotor to the rescue in Europe and perhaps in the US a year or two from now?
- 100K: That’s how many BYD EVs could be sold to Uber drivers in the EU and LATAM over the next few years. Uber and BYD have finalized multi-country preferred pricing for Uber drivers to encourage them to make the switch from petrol to clean energy.
I’d heard about discussions between the two parties several years ago – because a few entities asked for our help, but timing wasn’t right back then. I guess it is now. They’ll also work on intelligent driving features for the BYD vehicles.
There have to be a few OEMs kicking themselves for not being more open to working directly with Uber on these types of initiatives. Uber isn’t going away, and my bet is that they become an even more important player in the global mobility space in coming years.
_________________
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 global automotive and mobility sectors. We also provide a point of view that we hope educates and sparks debate.
The Sino Auto Insights team