Autonomous Vehicle ETFs

The recent SPACs of Aurora and Embark, on top of all of the other lidar and EV and eVTOL and AV companies that have gone public over the last year, got me wondering what the financial returns might be on a basket of these stocks.

And as soon as I wondered that, I suspected Wall Street had already created such a basket.

The answer turns out to be, sort of. There are a handful of ETFs that are roughly in the “next generation mobility” space, but a big confounder seems to be exposure to Tesla stock. Tesla stock has just been so successful over the last 12 months (last 10 years, really), that it seems that the performance of these ETFs must be dominated by how much Tesla stock they hold.

Indeed, in the chart above, you can see purple Tesla ticker is just way beyond every other option, whereas the Russell 3000 (up 33% in one of its best years ever!) lags the pack. Probably the Russell 3000 is a lot less exposed to Tesla than the future mobility ETFs.

Here is a different graph that plots some of the individual new mobility stocks over the last six months. You can see it’s kind of a mixed bag, with Tesla and NIO, the only two consumer-facing companies on the chart, performing quite well.

Seems like this should almost go without saying, but I’ll say it. None of this is investment advice. My personal financial portfolio consists of none of these stocks, but rather is almost entirely built on low-cost Vanguard Target Retirement Funds.

This is all just an interesting curiosity to me.

Hello South Bay

Cruise is coming to the South Bay! A new office will open at 840 W California in Sunnyvale, right next to downtown, the Caltrain, and Walmart Labs.

This is a big win for a lot of my former Voyage colleagues who are now Cruise colleagues. Voyage had a South Bay orientation, so returning to the office in 2022 was going to mean a long haul to San Francisco, for some of those folks.

The Mercury News reports that Cruise will have 400 employees in Sunnyvale, mostly engineering, so obviously this is about a lot more than just my former Voyage colleagues. Lots of Cruisers live in the South Bay, as indeed does most of the Silicon Valley automotive industry.

If you live in the South Bay and are interested in working for Cruise, email me your CV!

Under Cruise’s Hood

Last week Cruise conducted a 2.5 hour deep-dive into our technical architecture, with a particular focus our AV stack. The event was called Under The Hood, and now it’s available on YouTube for the whole world to watch.

You’ll see everything from how our perception stack operates, to planning, to the upcoming Cruise Origin, to how we’re going to roll out a ride-hailing service in San Francisco.

This is about as much information as any AV organization has ever put out about how self-driving cars work.

Check it out!

Mark Fields, Former Ford CEO, Buys 100,000 Teslas For Hertz

Hertz appoints ex-Ford boss Mark Fields as interim CEO | Auto Remarketing

Mark Fields just purchased 100,000 Teslas for Hertz’s rental car fleet. The deal is being hailed as biggest rental car purchase ever.

Mark Fields was CEO of Ford Motor Company during my time as an engineer there. I even got to sit in the background when Jim Cramer came to film a Mad Money clip with Fields.

Not long after I left Ford, however, Fields was unceremoniously shown the door, presumably the result of a languishing stock price. The stock price only got worse from there, while Fields moved into a new job in private equity.

Now Ford stock is roaring back, and Fields has a new new job, as CEO of Hertz.

The rental car company went bankrupt over the COVID summer of 2020, and is itself climbing, maybe quite “roaring”, back.

Bloomberg reports that Hertz is paying something close to full price for the fleet of Teslas, and the order size represents around 10% of Tesla’s annual manufacturing capacity.

This strikes me as a bet on Tesla’s status as an iconic car company. There are lots of nice cars Hertz could have purchased, but really only a few that I think most consumers would views as “iconic”: Tesla, Rolls-Royce, Mini, maybe the cute little electric Fiats.

I get why Fields didn’t purchase 100,000 Ford Mustang Mach-Es, even if those might be better cars.

The purchase also moves Hertz a bit closer toward Mobility as a Service. Reporting also mentions that Hertz will build a charging network to support its electrifying rental fleet and to complement the existing Tesla super-charger network.

Indy Autonomous Challenge

The Indianapolis Motor Speedway recently hosted teams of university students who programmed cars to drive the track autonomously, in an event called the Indy Autonomous Challenge. A team from the Technical University of Munich (TUM), in Germany, took first place.

This is such a neat endeavor, and the team that organized the event deserves a ton of credit for pulling it off, as do the university students, of course.

This is a somewhat different endeavor from building a regular self-driving car, just as race car driving is different from regular driving.

For one thing, this event was a time trial – each car was on the track by itself, and the goal was simply to complete two laps as fast as possible.

This eliminates some of the core responsibilities of autonomous vehicles, such as perception, prediction, and behavior.

Instead, teams can focus on pushing the limits of the vehicles’ localization, motion planning, and control software. Those limits turn our to be very, very fast.

It’s a lot of fun to watch.

NVIDIA DriveWorks 4.0

NVIDIA has been plugging away at self-driving for years now, and by all accounts making a lot of progress. They are arguably the world’s leading artificial intelligence company, due to the importance of its GPU products for deep learning.

Their self-driving business model isn’t totally clear – whether they will sell systems to OEMs or Tier 1 suppliers or even launch their own autonomous service. But they publish a lot of great research and software that impacts in the industry.

They just announced DriveWorks 4.0.

NVIDIA DriveWorks provides middleware functions on top of NVIDIA DRIVE OS that are fundamental to autonomous vehicle development. These consist of the sensor abstraction layer (SAL) and sensor plugins, data recorder, vehicle I/O support, and a deep neural network (DNN) framework. It’s modular, open, and designed to be compliant with automotive industry software standards.”

The closest analogy is probably Baidu’s Apollo self-driving ecosystem, which has seen more deployment within China.

Both of these are ambitious but somewhat under-the-radar efforts that I think may wind up surprising us.

Motional Passenger Experience

Motional has a neat blog post that describes several subtle features they’re incorporating to improve the passenger experience.

One theme I notice is trying to communicate features to users both in the mobile app and also on the vehicle. Different passengers will presumably respond better to cues in different places.

For example, the Motional app will work with the vehicle to make sure each passenger gets in the correct car.

“If a rider can’t find their vehicle, we’re exploring ways to help them locate their ride, such as using their phone to tell the robotaxi to flash its lights or honk its horn. Passengers may also be able to locate their ride by matching an identifier on their phone — a number, for example — to the same identifier shown on an LED panel on the vehicle’s exterior.”

Motional is also offering passengers several buttons that call a human customer support representative.

Importantly, the vehicles will be accessible to a wide range of passengers, through features like Braille, raised shapes, and physical buttons to complement touch screens.

There’s lots of other human-centered design features and ideas in the post, as well. Worth a read!

Aurora As A Service

Aurora Horizon and Aurora Connect, Aurora's subscription driver-as-a-service products will enable customers to deploy Aurora-powered trucks and ride-hailing passenger vehicles simply and seamlessly. (Photo: Aurora)

Aurora recently announced two programs – Horizon and Connect – to allow partners to rent driver systems for trucks and automobiles, respectively.

The announcement is pretty high-level and thin on details. There’s not a lot of context on how, exactly, the driver gets to the customer. Does the driver come with an Aurora vehicle, or can an Aurora driver be installed in a customer-owned vehicle?

But it’s a small step toward the dream of autonomy-as-a-service. The goal is something like Amazon Web Services, whereby any person or business with a credit card can rent a small vehicle, and then a small collection of vehicles, and then progressively larger vehicles that scale up and down, on-demand.