Waymo Has A First-To-Market Advantage In Autonomous Ride Services

Summary

• Autonomous vehicles will be launched within a year to provide autonomous ride services (ARS).
• It now is clear that Waymo (Google) will be the first major company to bring this service to the market.
• Waymo will have substantial first-to-market competitive advantages that will enable it to achieve significant market share in what is expected to be a very large new market.
• The ARS market in the US is projected to be $100 billion by 2025 and more than $750 billion by 2030.
• This has the potential of eventually valuing Waymo as a standalone business at as much or more than Google today.

I've been writing articles about autonomous vehicles on Seeking Alpha for almost two years, and I continue to believe that 2021 will be the year for the introduction of autonomous vehicles (AVs). It is now very clear that the first market opportunity for AVs will be Autonomous Ride Services (ARS) and not AVs sold at retail. It is now also clear that Waymo (NASDAQ:GOOG) (NASDAQ:GOOGL) will be the first to market and can establish substantial competitive advantages. In this article, I'll explain why ARS is coming within a year, refresh my projections of the ARS market, explain Waymo's strategy, highlight the strategies of potential competitors, and consider the significance of its first-to-market competitive advantages. First, let's review autonomous ride services and its market opportunity.

Autonomous Ride Services

Autonomous ride services is essentially Uber without a driver. It will be provided by fleets of autonomous vehicles based in selected metropolitan areas that offer trips to certain destinations along predefined routes. The autonomous driving technology used by these vehicles is based on LIDAR and high-definition maps that position the vehicle precisely on the map. Initially, the destinations and routes may be somewhat limited, but those will increase rapidly. A passenger will summon an autonomous ride using an app similar to Uber (UBER) or Lyft (LYFT), and the app will identify if that route is available.

The cost advantages of ARS over ridesharing are significant. In a previous article, I explained these cost differences: Uber Must Have Autonomous Ride Services. It's the classic case of a technology/capital-intensive business model replacing a labor-intensive one. Ridesharing companies like Uber and Lyft struggle to achieve breakeven on a typical trip at about $16, but ARS will be very profitable at about half that price.

Almost all companies developing autonomous vehicles have shifted their strategies to focus initially on developing autonomous vehicles for fleets providing autonomous ride services, instead of developing autonomous vehicles for the retail market. Most of these also intend to introduce their own autonomous ride services instead of selling fleets. The introduction of autonomous vehicles sold at retail has been delayed until 2025 or later.

The reasons that ARS will be the first market for autonomous vehicles are now obvious. The first generation of autonomous driving technology will be done using high-definition maps to place a vehicle precisely in its surroundings. This limits these vehicles to predefined routes and locations, which is perfectly fine for ARS. The app used to request an autonomous ride will be able to identify if the route can be served. ARS won't serve all routes, but will serve the 80%-90% most popular routes, and that works well.

ARS will be provided by fleets, most likely several thousand autonomous vehicles in each fleet, each serving a specific metropolitan area. This fleet-based strategy enables more control and increased safety. A fleet operations center will maintain the vehicles, work with local municipal authorities, and communicate with passengers when necessary. In the first stage (2021-2025), ARS will focus on selected municipalities, most like those in the Southern and Western parts of the country where the weather is favorable, road systems are easier, and local governments permit autonomous vehicles. Again, these favors ARS because ramping up these services will be capital intensive and require time to map local areas and initiate local fleet operations centers.

So, why do I believe that autonomous ride services are coming within a year? I realize that there are many skeptics who say that it will be a decade or more. First, the testing of autonomous vehicles has accelerated dramatically and has proven to be safe. For example, Waymo has driven its autonomous vehicles for 20 million miles without any significant accidents. Thirty-five companies did testing in California last year, driving autonomous vehicles 2.8 million miles, a 40% increase over 2018, without any significant accidents. Several companies are testing their autonomous ride services with actual passengers requesting rides, although they have a safety driver on board.

Secondly, there is a large pipeline of vehicles being prepared for autonomous ride services. Waymo has 80,000 vehicles in its pipeline, and Ford (F) and GM (GM) have stated their intentions to build tens of thousands by 2021.

Finally, Waymo is already providing autonomous rides to a limited base of customers without a safety driver on board. (See Video) I expect that it will expand this service this year and fully launch it in the Scottsdale area by 2021, if not sooner. GM has indicated that it will introduce its ARS in San Francisco next year, and Ford has recommitted to having its autonomous ride service available next year also.

Autonomous Ride Services Market

ARS will be a significant new market, perhaps one of the largest new markets of this century. This is why companies are investing heavily. Let's look at some updated forecasts.

The projection is 30,000 autonomous ride service vehicles by 2021, generating more than $4.5 billion in revenue. For those of you interested in the math, here it is. Each ARS vehicle is projected to generate approximately $150,000 in revenue per year, based on 50 paid trips per day at about $8.75 per trip, operating 350 days in a year.

$8.75 per 7-mile trip is about half of the price of ridesharing today, so ARS can quickly take market share. ARS can be profitable at this price. I estimate profit contribution at about $2.25 per trip, while ridesharing is still not profitable.

These 30,000 ARS vehicles will generate about $4.5 billion in revenue. To put this into perspective, this equates to an insignificant percentage of total miles driven (about 0.1%). The ARS market will grow rapidly as more fleets are introduced into new municipal areas and the number of autonomous vehicles in each fleet increases. By the end of Stage 1 in 2025, approximately 750,000 autonomous vehicles will be in use to provide ARS, generating almost $100 billion in revenue. At this point, ARS will seriously cannibalize the ridesharing market and provide about 2% of the miles driven in the US, without expanding beyond more than 30-50 metropolitan areas in Southern and Western part of the country.

These limited metropolitan areas have the potential to provide a sufficient market for autonomous ride services in the first stage through 2025. For example, the greater Phoenix metropolitan area could support 100,000 autonomous ride services vehicles by providing only 7.5% of the miles traveled. Other metropolitan areas in Florida such as Miami, Collier County, and Orlando can support even more. Municipalities in California and Texas will enable many more metropolitan markets.

Waymo's Strategy

Waymo, Google at the time and now Alphabet was one of the first to work on autonomous vehicle technology. It has invested more and has made more progress than others. Waymo's autonomous vehicles have driven 20 million miles on public roads in 25 cities. Additionally, it has "driven" tens of billions of miles through computer simulations. Just think about that for a minute: 20 million miles of autonomous driving with no serious accidents. It's ready to release its autonomous vehicles.

Waymo is using 600 Fiat Chrysler (FCAU) Pacifica hybrid minivans for its testing. It has also announced that it plans to acquire 62,000 more and also purchase 20,000 Jaguar I-PACE electric vehicles. It built a factory, along with Magna, in Southeast Michigan to retrofit these vehicles to be autonomous. If Waymo puts these 82,000 autonomous vehicles into service by 2023, it would represent more than 50% of the forecast in the table above.

The Waymo One app enables riders to request a pick-up and drop-off location, similar to Uber and Lyft. If it is a route that it serves, then it provides an autonomous ride. Waymo has been cautious in testing its ride services. It launched an early rider program in April 2017 for 400 people in Arizona, serving a 100-square-mile area. These rides were autonomous but had a safety driver on board. Waymo is increasing the number of people served and the area covered. At the end of last year, Waymo started providing autonomous rides without a safety driver.

Waymo has also continued testing in California. In 2019, it had 148 autonomous vehicles that drove almost 1.5 million miles without any serious accidents. It recorded about two dozen incident reports, but these were almost all caused by other vehicles at very low speeds, with only minor damage.

Waymo has also begun targeting other municipal areas. In September 2019, it announced that it would start testing and launching an ARS in Florida, beginning with Miami but also adding routes to Fort Meyers and Orlando. Then, it announced that it was developing detailed ARS maps for Los Angeles.

Waymo's ARS Competitors

Even though Waymo will have a first-to-market advantage, there will be a lot of competition. ARS is a very large market opportunity, and other companies also are targeting it. Let's look briefly at some of the other expected competitors for the US market and their strategies.

I currently rank GM Cruise Automation as the second company behind Waymo. It intends to launch its autonomous ride services in San Francisco. Cruise currently has a fleet of more than 200 autonomous vehicles in San Francisco, using its Cruise Anywhere app to schedule rides. In 2019, its autonomous vehicles drove more than 800,000 miles in California. Cruise recently unveiled the Cruise Origin, a six-person taxi purpose-built for autonomous ride services in urban areas such as San Francisco. Cruise is expected to launch its ARS in 2021. SoftBank and Honda have invested in Cruise Automation, giving it a value of approximately $15 billion (almost a third of GM's valuation). Most likely, Cruise will be spun off as a separate company.

Ford Autonomous Vehicles LLC has made it clear that its initial strategy is focused on ARS. It has been testing its AVs in Austin, Miami, and Washington DC. Some reports claim that Ford plans to build 100,000 ARS vehicles starting in 2021. Its original strategy was to provide AVs to companies who would provide ARS, but that changed when the major players in the industry realized they needed to have control over their customers. Unlike most other competitors, Ford expects to use hybrid autonomous vehicles instead of fully electric. It believes this approach works best because the autonomous platform drains battery life and AVs for ARS are more profitable with less downtime for charging.

Aptiv (APTV), now a joint venture with Hyundai Motor Group (OTCPK:HYMLF), uses technology based on its 2017 acquisition of nuTonomy. In May of 2018, Aptiv launched its ARS service in Las Vegas with a fleet of 30 BMWs in conjunction with Lyft. Now, that fleet has completed over 100,000 autonomous rides with 75 BMW 5 Series sedans.

Uber's strategy is to build a very large, although most likely unprofitable, ridesharing business and then convert its customer base to a profitable autonomous ride services business. ARS will cannibalize ridesharing because of its highly favorable economic advantages. So, Uber has no choice but to move to ARS. Unfortunately, Uber is in an unsettling position for ARS. It is developing its own autonomous driving technology but suffered a major setback with a tragic fatal accident on March 18, 2018. It then suspended all testing but recently restarted its program. Uber's initial target markets for ARS appear to be Dallas and Washington DC, in addition to its testing in Pittsburg.

Like Uber, Lyft is at risk of cannibalization from ARS. Lyft's ARS strategy is different from Uber's in that it has a two-pronged approach: both developing its own technology and partnering with others. It has preliminary pilot partnerships with Waymo and Aptiv, but it remains to be seen how these partnerships will work out since everyone wants to control the customer. Lyft's investment in developing its own autonomous technology doesn't seem up to many of the others. Lyft started testing its AVs in California last year with 19 autonomous vehicles, which drove about 43,000 miles.

Apple (AAPL) is the wildcard in this market. Apple has been very secretive about its autonomous vehicle program, code-named Project Titan, but as I've written previously (Deciphering Apple's Autonomous Vehicle Strategy), there is a lot of evidence that it will enter the ARS market. It has invested significantly and has filed hundreds of autonomous-driving patents. At the end of 2018, as part of legal action against a former employee, Apple claimed that it had 5,000 employees who were disclosed on its autonomous driving project. Most likely, Apple will be a follower, letting other companies create the market before it enters, and then come into the market with a high-quality service and second-generation autonomous vehicles. I now project that Apple will be a 2023 entrant into the ARS market.

Finally, I'm sure many people are wondering about Tesla's (TSLA) promise of the Tesla Network, a fleet of customer-owned and company-owned Teslas to provide autonomous ride services. Even though I own a Tesla and think it is the best semi-autonomous vehicle, its autonomous technology platform works well for highway driving, but it is not yet viable for municipal driving where frequent turns are required. Also, the Tesla Network business model for providing swarms of autonomous vehicles for ride services is not workable.

Waymo's First-to-Market Advantage

The importance of a first-to-market competitive advantage varies. It can be substantial in some cases but not worth the early risk in others. With ARS, Waymo will have substantial first-to-market competitive advantages.

The initial obvious first-to-market advantage is the development of autonomous driving technology, primarily software. Autonomous driving software is very complex and takes time to develop and test. Waymo started development earlier and has invested more than other competitors. It has been testing its software longer and much more extensively. Waymo's 20 million miles of testing on public roads is well ahead of any others. While competitors may be able to learn a little from observing Waymo's AVs in action, there isn't much they can do to reduce the software-development cycle, short of stealing it (which Uber tried to do).

The second advantage, and maybe the most important one, is that ARS is a regional, not global, market with services provided in selected municipalities. This is very different from other high-tech products. Once a competitor enters a specific municipal market and captures a significant major share, others will be discouraged from entering the same market and trying to displace the incumbent. Waymo appears poised right now to enter several of the most lucrative municipal markets and could secure these markets before others even get started.

Ride-request and dispatch, as well as fleet management, are important operational functions for autonomous ride services. These are often taken for granted, but they are difficult to develop and perfect the customer experience. Here again, Waymo has a first-to-market advantage with its Waymo One app and its extensive customer experience in Arizona. Competitors, except possibly Uber and Lyft because they already have ridesharing apps, are behind in these operational functions.

Waymo depends on others to manufacture its basic vehicles, but its strategy compensates for this deficiency. It has ordered 80,000 basic vehicles from Fiat Chrysler and Jaguar to customize in its factory in Michigan. This enables it to get the first-generation of ARS autonomous vehicles to market faster, but this could be a competitive disadvantage when the ARS market shifts to second-generation autonomous vehicles designed specifically for autonomous ride services without any driver controls.

Experience really matters in this market. With more experience, a company's ARS will be safer and able to cover more unique situations. Also, it can scale the operations side of the ARS business more effectively.

There are always some risks to being the first to market. In this case, it's the risk of a fatal accident. While AVs will be much safer than human drivers, there still will be some accidents, and the early ones will be overblown by the press and those who are fighting AVs.

Waymo's Potential Market Share and Revenue

With its first-to-market advantage, Waymo could achieve significant market share in Stage 1 of the ARS market. Assuming it deploys the 80,000 autonomous vehicles by 2023, it would have about 50% of a projected $23 billion market. While that is not significant relative to Alphabet's total revenue, it could be a very valuable asset. At about $12 billion in revenue in 2023 and growing rapidly, Waymo would be similar in size to Uber today, which is currently valued at $60 billion. However, Waymo should be very profitable, and Uber is currently losing money.

By 2025, Waymo's share of the ARS market will most likely decline as competitors expand their fleets and enter into more municipal markets. With a 35% to 40% market share, its revenues could be $35-40 billion by 2025.

The ARS market will continue to grow rapidly through Stage 2, which goes through 2030. My projections are that the market could reach $750 billion. By this time, there could be 5 million ARS autonomous vehicles throughout the US that account for up to 15% of the total miles driven. Waymo could still be holding a 25-30% market share, giving it revenue of approximately $200 billion.

At some point during this growth, it would not be surprising for Alphabet to spin off Waymo as an independent company. This seems to be the trend with GM and Ford doing this as well as others. It makes sense because ARS is a very different business, and it will require a lot of capital.

I won't speculate on forecasting its potential valuation, but as the market leader with barriers to entry in a huge rapidly growing market, it could possibly be worth as much as the rest of Google.

Analyst’s Disclosure: I am/we are long GOOG, AAPL, GM. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.

Comments

[Ian Humble]

I don't know the full details of how road maintenance is paid for in the US (I gather its a mixture of general taxation and fuel duty), but how will ARS profits being sent out of town at the expense of local drivers be seen by the locals? Not favourably I imagine. So will additional taxation will be added to make a level playing field, reducing their viability further ?

[Michael McGrath]

Good question. Much of it today comes from the gasoline tax, and that is already in jeopardy with electric vehicles. More municipalities are taxing Uber and Lyft now. I would expect that to continue with autonomous ride service vehicles. And it would be fair to do that. While such a tax may slow adoption a little, I don't expect it will change the viability.

[User 50844673]

About Regulatory interests and testing regimes that are coming faster than you can say "Hoi, Taxi!"

Moneyball @DKurac
Mar 10
#China to open its 1st #L4 autonomous driving testing facility in Chongqing, local media says.
180 km testing facility to be made by Baidu Apollo & local gov. It will be able to support >100 vehicle tests. Jun start, the equipment to be installed. Jul start, trial runs to start. twitter.com/...

[rog4024]

"Offer trips to certain destinations along predetermined routes." Isn't that what a bus does and a lot cheaper?

[Tdot]

Pretty much, except the driver and steering wheel and door opening lever and such.

But yes, the current plans for the "first generation" AVs are mainly as taxi cabs and shuttle buses operating on carefully mapped and geofenced routes, in other words SAE Level 4 Autonomous.

SAE Level 5 is thought to be a decade away from full implementation and being "consumer ready", regardless of Musk's personal beliefs and hopes on the matter. Remember, Musk said in 2016 that Teslas would be be able to drive themselves from coast to coast with no human occupants, and deliver themselves to customers.

Ford and GM seem to be setting up for SAE Level 4 operations in 2021, with more and more "public" demonstrations coming in the next few to several months.

Meanwhile Teslas are still just trying to learn how to avoid hitting things and going off the paved road in semi-AutoPilot driver assist mode, and in the valet self-parking app that sort-of released late last year to much fanfare and laughter as the self-parking Teslas showed the maneuvering and parking skills of a 5 year old child trying to see over the dashboard while struggling to reach the brake pedal.

[davel]

Your timelines seem wildly aggressive from what I have read.

Waymo does not have first mover advantage. It does not possess a self driving car. What it possesses is a taxi service like what has been offered for 100 years.

Everyone seems to think waymo is so far ahead. Maybe they are. Maybe they publish data because they were spun off and need a story to tell. Google was first with google glasses. They rolled out google fiber, their cable service years ago. How are those initiatives doing? Google does a lot of stuff. Some very successful and others not.

Waymo ride service was offered about a year ago. As I understand it the service is essentially a taxi service like Uber and lyft. This has not changed.

You make projections of how many automated cars will be available. How much does each car cost? And with the limitations you impose will people choose this service or go with a comparable service that gets them to where they are going when they want? In short I question if it would be profitable in a few years and if the service companies would be comfortable with the capital risk they incur for a marginal improvement in one of the dimensions.

[Michael McGrath]

Some good questions. Waymo has developed the autonomous driving technology and it is putting that into 80,000 base vehicles that it acquired from Jaguar and Chrysler. It will use these to provide driverless taxi service with significantly better economics than Uber or lift because it eliminates the cost of the driver. Waymo has been testing/offering ride services like you said for more than a year. These have had a safety driver on board until now. This is a big difference. Autonomous ride services can be profitable at half the price of Uber, so people will begin to use it where available. If you are interested, I explain all of this with the economic models in my book -- Autonomous Vehicles: Opportunities, strategies, and Disruptions. Thank you.

[davel]

@Michael McGrath

As I understand it they have not acquired 80,000 vehicles. I see it as an order of intent. Let us say they “purchased” said cars. Where will they deploy them? That comes to laying out $4B for a fleet of cars they are not using if you assume each car costs $50k. That is a lot of coin to spend on vehicles that will sit in a garage. This of course does not include the additional technology they need to introduce to each vehicle to support autonomous driving.

[User 50844673]

@Michael McGrath @davel

"As I understand it they have not acquired 80,000 vehicles. .... laying out $4B for a fleet of cars they are not using if you assume each car costs $50k....does not include the additional technology they need"

Michael said elsewhere in comments they'll cost circa $100K each - $50K for the technology aspect - so it's $8 billion. Much much more than it costs to put a Licenced Taxi or Uber/Lyft car on the road. The Insurance will not be less than a Licenced Taxi is charged - specially because it is remotely Autonomous - unproven - with no driver on board.

3 million ARS vehicles spread across several operators Waymo, Telsa, Cruise Mobileye etc may cost operators a mere $300 Billion just to build them. Then comes the costs of creating an Uber like Rideshare IT system and administration.

Then they need to PAY THE ACCUMULATED COSTS (while making a massive loss at startup) to actively COMPETE for YEARS against with all the existing competition of Taxis and Rideshare and other Public/Private Transport providers who currently hold the 100% of Market.

@Michael McGrath "significantly better economics than Uber or lift because it eliminates the cost of the driver. "

They DO NOT eliminate all the costs carried by the Owners of rideshare vehicles. Drivers only get "paid" whatever is left over AFTER they provide the vehicle, operate it, repair it, insure it, get accident repairs done, replace tyres, pay govt licences and fees, iow pay all the costs involved in providing the service as a owner/driver PLUS operating the car.

The savings of driver Labour costs is minimal - maybe half at best the PITTANCE Uber pays the driver. Out of that payment then the owner needs to cover the Asset value loss from the 60K-100K miles per year -- a cost which is significantly higher than an Depreciation charge to their small business.

Most Taxis/Rideshare vehicles are worthless after 3 years full time on the road and handling paying Passengers!!!

By dumping the "owner/driver model" Waymo et al will be immediately picking up ALL the costs and chargers that Uber corp et al have NEVER had to pay in the first place.

Taxi driver who driver cars owned by the taxi business itself are paid ~40% of the fare. They can afford to do this because Fares are double Uber's rate!!!

A Taxi business is using much cheaper cars than an ARS network would.

IF Taxi drivers were unpaid slaves - and Taxis rates were what Uber charges - the Taxi Owners and Network Companies would go Bankrupt and shutdown within weeks!

Charging Fares Half the Price of Uber rideshare is not Viable - they'd be out of business on the first day - unless sitting on $BILLIONS of Investors CASH to BURN.

These "off the top of my head" estimates and claims and pronouncement DO NOT ADD UP for an ARS ride-sharing Business at all. The claims are ILLOGICAL and UNSUSTAINABLE on first Principles.

Unfortunately the Hyped Spin never stops - @Michael McGrath "This is a big difference. Autonomous ride services can be profitable at HALF THE PRICE of Uber "

THAT is living in fantasyland stuff without a shred of evidence provided to support the CLAIM. This is very poor form by an INVESTMENT ADVISOR Author on SA. @Michael McGrath you are not a 'throw away' anonymous commenter here.

Enough of the "read my book" excuse Michael - Please SHOW HOW an ARS Rideshare/Taxi service can set prices 50% BELOW UBER'S MILEAGE RATES and the OTHER Charges to customers and survive as business longer than a day.

Investors deserve nothing less than that - given the repeated claims and predictions about ARS Glory. TY

It's not asking too much either. It should be a GIVEN here by investment advisors and self-appointed "industry specialists" who have written books on the subject.

Right?

[johnafish]

I don't see it, the*only* advantage is there's no driver to pay yet the disadvantages and additional costs are significant, these include the extra cost of the vehicle, the higher cost of insurance, the likely higher cost of cleaning and repairing damage. The customer is inconvenienced by route and weather limitations and without a doubt these vehicles will encounter situations they can't deal with. So the customer must have an incentive to use the less capable driverless service which is another 'cost'.

I think whoever did the math on driverless cars is mistaken, I expect it'll be less profitable than other ride sharing service and that should worry investors.

[Michael McGrath]

I did the math in a previous Seeking Alpha article: seekingalpha.com/...

Basically for a typical $16 trip, Uber pays the driver $12.25 on average, including the use of his vehicle. The estimated cost of an autonomous vehicle including everything (depreciation, fuel, insurance, cleaning, and maintenance) is estimated at about $52,000 per year. For 30-50 trips a day over 250 days per year, the cost is about $2.50 - 3.00 per trip. So, it's much cheaper. Route and weather limitations will initially be managed by telling the customer it isn't available, so the customer will use Uber or Lyft instead for that trip. I forecast that the price for ARS will be about half the price for ridesharing and still be profitable.

I fully describe this in my book (Autonomous Vehicles: Opportunities, Strategies, and Disruptions). Three years ago, I was skeptical too, but then I did the research. It's basically technology and capital replacing a labor-intensive business model. That has proven successful in most cases historically. Thanks.

[johnafish]

I bet someone also calculated Uber's profits for their business plan. I wonder how that compares with reality.

I call nonsense on the whole robotaxi industry from a business perspective, technically interesting though.

Perhaps it'll make money with tourists as a novelty at a ridiculous price, a bit like Concorde did in it's later years.

[User 50844673]

@Michael McGrath " For 30-50 trips a day over 250 days per year, the cost is about $2.50 - 3.00 per trip. "

"I forecast that the price for ARS will be about half the price for ridesharing and still be profitable. "

Sorry that's simply BS fantasy.

You have the option to SHOW your work though. The devil is in the details.

[J20]

What do you think about ground penetrating radar? Do you think this is how they will deal with snow?

[Michael McGrath]

It's still in the research stage now, but I think it's fascinating. Right now the practice is that AVs will drive like humans in the snow by looking for physical landmarks like trees, signs, etc. and then estimating where the road is, unless there are other cars or tracks to follow.

If ground-penetrating radar actually works, then it would be much more accurate than humans. From what I've seen so far though, there are still obstacles to making it work reliably. Thanks.

[Tdot]

Ford is using LiDAR and known (geomapped) roadside features to find the road in snowy conditions - media.ford.com/...

"LiDAR, on the other hand, is much more accurate than GPS – identifying the Fusion Hybrid’s lane location right down to the centimeter. LiDAR emits short pulses of laser light to precisely allow the vehicle to create a real-time, high-definition 3D image of what’s around it."

"An autonomous vehicle creates the maps while driving the test environment in favorable weather, with technologies automatically annotating features like traffic signs, trees and buildings. When the vehicle can’t see the ground, it detects above-ground landmarks to pinpoint itself on the map, and then subsequently uses the map to drive successfully in inclement conditions."

[jeanchartman0587]

canon mx922 b200 error

[daveruns2win]

I think the market for these services can be segmented and different strategies will work better in different situations. For instance, owning an autonomous vehicle and sending it to pick up the kids after soccer practice, taking grandpa to the doctor, or simply meeting you at the airport on your return from a business trip makes a lot of sense for an individual owner. It also cuts into the common reasons that many people use current ride sharing services. I can also imagine a business owning a car to shuttle employees and visitors to airports, hotels, etc. Pizza restaurants may want to own their own cars for branding and quality control. I can also imagine new services, where you order online and send the car to pickup your stuff. I think the convenience of having your own autonomous car will be important in more affluent areas and suburban areas where parking is not an issue. But obviously this will not work as well in large cities with limited expensive parking. It is also not the most efficient use of the technology, but other benefits often trump cost.

[Michael McGrath]

Very perceptive. The market will be segmented like this. I cover all of those cases in my book -- Autonomous Vehicles: Opportunities, Strategies, and Disruptions.

[radri]

Michael,
Very thorough article on ARS, AV and Waymo...very informative, thanks.
Was wondering if you’ve done any research on Luminar and their LIDAR product?
Rob A

[Michael McGrath]

Rob

I haven't done very much research on specific Lidar products. Initially, I thought that Lidar would be provided by Luminar and other Lidar companies. But it now appears that many of the companies developing AVs are also developing their own Lidar since it is a critical technology.
Michael

[Tdot]

It isn't so much that they are "developing their own LIDAR", it is more that they are developing their own ways of integrating LIDAR into their AV designs so they don't look so much like spinning soup cans sprouting from the roof like some kind of weird parasitic fungus.

[User 50844673]

Lois, have you already read and absorbed (or even know about) this Dutch Safety Board Report on ADAS and autonomy from Nov 2019?
Who is in control? www.onderzoeksraad.nl/...

Using technology to improve safety of road suers is a great idea . But it has it's limitations too. Corporate driven Hyped PR does not amount to genuine facts nor evidence of a social good.

But who is going to pay for the "Black Boxes" that can withstand a Tesla battery fire being placed in every single Robotaxi vehicle ... and the resources to analyses the data every time there's an accident injury or death involved? And what about the Govt approved tamper proof video recording installed internally of each and every vehicle? That's not cheap either.

I've checked into Google scholar since 3 years ago, and am yet to find a single Paper that supports the need for and viability of autonomous vehicles and robotaxi services now or in the medium future. I know how to navigate that domain and even though I am not going to pay $30-$60 each to read the subscription papers I'm aware enough of how to parse what the overall consensus conclusions are thus far by objective credible academics and scientists. doing proper research and analysis of the issues involved at all kind of levels.

The overall message I get from these experts is that FSD Autonomy and Robotaxis are an impractical unviable expensive solution looking for a problem that doesn't even exist.

But what I really want to know is what happens to the drunk client who falls dead asleep in the back seat of the Robotaxi once they arrive home in the middle of now where after they projectile vomit all mover the interior? Sure charge them to clean it up but who gets them out of that car? (smile)

[Rjzq]

None of your objections are related to autonomy. They’re about ride sharing in general and electric cars in general.

[User 50844673]

@jakexb
The article is about "autonomous-ride-services" the later "market" issues etc etc can't be excluded from the conversation.

My comments are very much directly related to Autonomous Driving (claims assertions and disinformation) in general and specifically.

You must have missed those parts above and down here seekingalpha.com/...

[Jack'sson]

It isn't fair to compare autonomous vehicles or TESLA batteries to perfection. AV will be safer than taking an Uber and TESLA's are the safest car on the road.

[S_Archer]

@Michael McGrath
How much does a L4 car cost today and next year?

[Michael McGrath]

I use an estimate of $100K in my book (Autonomous Vehicles: Opportunities, Strategies, and Disruptions). Roughly, this is a high-end base vehicle at $50K (one with ADAS and other similar features removed because the autonomous systems replace these) and another $50K for the autonomous system. These costs will continue to go down, but in the economic model of ARS, it doesn't matter that much since the utilization rates are so high. It's more profitable to have a $100K vehicle used 50% of the time than a $60K vehicle used 5% of the time.

[clcellve]

@Michael McGrath
Your point on the relation between profitability and utilization rate is a point many miss. Ride share vehicles will have a high use rate - they're not sitting in a garage or parking lot - they're out making money, mile after mile. And for electric vehicles, the maintenance costs are significantly less than an internal combustion engine. They're rolling money machines....

[User 50844673]

I use an estimate
the economic model of ARS,
since the utilization rates are so high

Credible Objective Peer-Reviewed Evidence?

I guess I'll have to read the book. :)

[User 50844673]

No mention of Mobileye ARS in China in 2021?

"Ridesharing companies like Uber and Lyft struggle to achieve breakeven on a typical trip at about $16"

They don't own a single vehicle nor a Smart Phone themselves - and they still cannot make a profit running an in-house basic remote IT APP system?

Their ONLY cost is their OWN Administration - that's it - and they are miles from making a profit. In many countries they are banned and will likely remain banned. ARS systems as well.

They don't pay for a vehicles running costs nor registration, don't pay for Smartphone or connectivity, or pay the driver licensing fees and the car insurance of the cars. The drivers own them (pay everything) and/or use cheapest rental/lease options packaged for Rideshare drivers.

If a 'full time' Driver's use their own vehicle the daily mileage is rapidly destroying their vehicle's asset value to next to zero in 3 years if they drive full time. They are being paid next to nothing (min. wage) per hour and mainly living off the $$$ paid by Uber which does not compensate them for the destruction of their Assets (or their payments to the Car Hire service) and the annual running costs PLUS their IncomeTax and GST/Sales Taxes.

Let alone their Labor Time during and outside their driving time..

Suddenly Waymo, Uber Lyft and Tesla are going to be able to OWN their own cars, pay for their running costs, Govt regulations fees, and make a profit on an average $16 a user trip?

While driving around (on average likely) 1000 miles for every 300 miles paid for by a customer? And they've been saying that will offer even cheaper fares to Rideshare users than they are at already - much much cheaper than a regular taxi per trip.

I believe these "business income/profit models" and marketing analysis are primarily based on extremely biased Theories - their assumptions are seriously flawed - imo - all being absolutely best case scenarios that totally unrealistic long term. But we'll see.

You can be an Uber driver in a $20K car. In India and elsewhere it's a $5K car.

ARS are going to make $100 Billion in Revenue in 2025 deploying 'remote controlled' IT APP systems running Autonomously equipped $50K to $80K 100% BEV vehicles and suddenly make a Profit and Compete with on price, service, efficiency, speed and safety of human drivers in taxis, rideshare and hire cars - in their dreams maybe.

[Louis Stevens]

@Muscone

You lack the fundamental understanding as to how these companies plan to govern the self-driving network.

No company in their right mind would field a fleet and expect to profit from it.

Uber does not plan to, nor does Tesla. I don't know what Alphabet will do, but no one should be buying Alphabet for self-driving.

[User 50844673]

Yes but I know 1+1 still equals 2.

And that follows that fundamental truth.

Feel free to fill me and billions of others on this planet about "how these companies plan to govern the self-driving network."

Because at present it's the world's most closely held secret. And what I have seen make absolutely no logical sense whatsoever.

The information in the article is (imho) a word salad festival that goes round in circles but says so very little of substance.

Follow this thought: Person A pushes phone APP button to order ARS transportation from A to B.

APP central finds closest ARS vehicle and directs it to the Customer.

Customer waits and watches progress and notes wait time before ARS vehicle arrives.

Vehcile arrives (eventually) person get in and ARS system drives ever so slowly to Destination B and stop ) somewhere.

Person get's out.

ARS IT APP System Charges fees to customers credit card.

Now tell me again HOW and HOW MUCH this ARS entity makes per Trip in Profit?

Uber and Lyft and OLA make multiples of $X in Losses and will for eyars ahead.

Until by some "AI miracle" ARS comes along costing I don't know $100 Billion $200 Billion or more in R&D (to be recovered in the next decade out of *Future Profits*) even before they have one car on the road and this *saves" all these currently broke and indebted Ride-share companies (and the dozens of NEW ARS ones claimed to be appearing any moment now) all becasue they OWN their own ARS vehicles in the MILLIONS and don;t have to pay a "driver".

A driver who really doesn't actually get even PAID PROPERLY or FAIRLY for the Services they provide and the Risks they take all on their own.

OK .. now shoot.

I'm all ears. Seriously. Help me understand this "brave new world" that at present only exists as a Theory in people's "minds" and whiuch they to date have been unable to explain to anyone adequately except using "metaphors" such as

*it's cheaper - better - safer - without a driver*

[Louis Stevens]

@Muscone

It's not a closely held secret at all... Watch interviews with Dara (Uber's CEO).

He clearly explains simply how they will roll it out. Literally in like two sentences.

[Johnathan21212]

Basic management:

FIRST MOVER IS NOT AN ADVANTAGE!
--> It is only an advantage if it is accompanied by other REAL competitive advantages such as
1) Economies of scale that can't be replicated via just having a ton of capital (Such as REAL network effects like Facebook/social media for example)
2) High switching costs
3) Highly differentiated products.

Why?
The first mover spends an enormous amount of capital and TIME to be the first to market that must inherently be made up once it gets to scale. The second/third movers face a FRACTION of the upfront costs because they just copy the new model.

[Michael McGrath]

I agree with you in general. In Waymo's case, it will be able to achieve the first two and maybe the third. Also, in this market, unlike many others, competitors can't just copy the product. They will need to develop their own software and gain their own experience to prove it works safely.

[Microcap Growth Investor]

Interesting to see Waymo raise $2.25B today from a group of non-strategic investors, like Silverlake and Canada Pension Fund. Maybe some strategics are still coming, but on its face, seems weird, given their huge cash hoard of over $110B in net cash at Google.

[Louis Stevens]

@Microcap Growth Investor

It's likely to ensure incentives are properly structured.

[Michael McGrath]

Yes, that was announced after my article, and it illustrates that fleet-based autonomous ride services will be capital intensive. A fleet of 5,000 AVs will cost more than a half a billion dollars. An eventual fleet of 1 million AVs could require a $100 billion investment. To me, Waymo taking on outside investors indicates that it is getting ready to build and launch its fleet. Waymo also indicated in its announcement that it may do an IPO at some time in the future. Google has a lot of cash but may not want to use a significant portion of it. Most likely it has done the numbers on capital investment and sees how much will be required.

[Microcap Growth Investor]

@Michael McGrath ,

Good article! A couple comments.

I realize that AV's are going to be very expensive, but 1m AVs at a cost $100B seems too high of a forecast? That would be $1m per unit, which would prohibit an ROI *IF* your revenue projection of $150k annually per unit doesn't pan out OR if profit margins aren't at least 15%-20%. Both of which seem high for Ride-sharing-given how competitive it will be. Its likely that competitors will drive margins down in ride-sharing (at least initially in the first 5 yrs or so).

Also, note: At Cruise's AV day when they showed off their new Cruise Origin AV (about two weeks ago) GM/Cruise said the Origin (including electric motors and batteries) would cost between $40,000- $50,000 per copy when produced at scale. Now, I realize that doesn't include the SDS software, data centers (since they will all be connected), and lidar etc... but these costs will go down at scale.

Additionally, although they are spending small fortunes on AV's (example Ford has committed over $4B by 2023) none of the OEM's have anywhere close to $100B to invest in AV's. Even if AV's start throwing off cash to reinvest to get scale, $100B for 1m AV's seems too high. We'll see...

Ps. My comment above was that it was interesting that Waymo ISN'T getting any strategic investors (OEM partners) as investors.... That is a small red flag for me. Why isn't Fiat Chrysler, Renault/Nissan or Tata (Jag) investing? They are all supplying Waymo with vehicles and are partners with Waymo. I think it likely means the OEMs are keeping their options open. The OEM's will just be supplying the "dumb" vehicles to Waymo that inserts the "brain" (SDS) and runs the high margin service. Waymo is positioned to capitalize on the most lucrative parts of the mobility stack and their OEM partners will be squeezed into the lowest value piece - supplying a vehicle. Why should the OEMs stick with Waymo given AVs will likely begin to cannabalize their existing business? In other words, OEMs that partner with WAYMO may limit their downside somewhat as the industry gets disrupted, but they won't really participate in much upside in the biggest opportunity to OEMs since the advent of the car!

[Risk Advisor]

Vooch--

With respect to Ford, who books the revenue and profits, Ford or Argo Al, (in which VW is also invested)? Does Ford just profit as a percent of its investment in Argo Al? In other words to what extent does it have control? Or is it similar to its joint ventures around the world such as China where Ford books only 50% and the Chinese book 50%?

[S_Archer]

So you are saying that there will be 30K Level 5 (no driver) cars on the road next year?

And there will be $4B in revenue from these NO DRIVER cars in 2021

and $100B in revenue from this in 2025

Seriously?

I assume you are off by a factor of 10,000 or 100 years.

[Michael McGrath]

These are Level 4, not Level 5. Level 4 enables the vehicle to drive itself under limited conditions such as geo-fencing. This enables autonomous ride services. Level 5, as you imply, is much further away.

[S_Archer]

L4 .... got it.
so $4B from L4 driverless cars in 2021? 30K L4 cars in 2021

there will not be 3000 L4 cars on the road as driverless rideshares in 2021.

How many L4 Driverless cars do you think are working as robotaxis today?

[petjumper]

Geo-fencing doesn't help to get to level 4 in cities. Sorry, but your entire reasoning is built on incorrect assumptions about the technology involved and the massive problems, even Waymo is still not able to solve as to inner city traffic.
Back to square one, please!

[clcellve]

I try to follow the topic, but I may have missed a few announcements....

When did GM say deployment in 2021? Last year, they said deployment in 2019, then postponed indefinitely, with recurring statements that things are going well, but I've seen no deployment update. I have seen statements that GM is in negotiations with the NHTSA on deployment/regulations specifics. I've seen no statement that Waymo is in similar/serous negotiations with the NHTSA.

GM spinning off their system is, as far as I know, sheer speculation. To my knowledge GM has not made any statements on the subject. To me, there is an excellent case to keep Cruise "in-house" to seamlessly integrate (insofar as possible) design, manufacture, and maintenance. This integration is a potential/probable competitive advantage for GM.

GM will be promoting their own driver assist system (Super-Cruise), in tandem with deployment of Super Cruise. And increased electric vehicles. In 2-3 years, GM will be a very different company.

Don't forget trucks - big money in long-haul autonomous truck transport - UPS, USPS, Walmart, Amazon, others, will be eager to save labor costs on highway trucking. And electric trucks, too. Big money for GM to offer driverless electric trucks.

To say Waymo has a first to market advantage assumes they will be first to market (evidence of that??), and then the question of scale - first to market in Scottsdale, or in several major US cities? We'll see.

[Michael McGrath]

Good questions/comments. GM has already given up much of the equity in Cruise by selling equity for more than $5 billion to SoftBank and Honda. Super Cruise is a very good semi-autonomous system. The NHTSA issue involves the elimination of the steering wheel and other driver controls, which GM is asking for. The NHTSA doesn't have jurisdiction over the need for a driver in a car. Those are state-level issues. Waymo's AVs have steering wheels, they just aren't used. GM did delay from 2019 but has implied 2020. I believe that its more like 2021. This isn't contingent on the NHTSA ruling. Yes, you are correct that trucking is a big opportunity. I cover those also in my book -- Autonomous Vehicles: Opportunities, Strategies, and Disruptions. Regarding Waymo being first, I tried to describe the reasons for that in this article, so I won't repeat them.

[x001464]

Double lol. I work in autonomous vehicles. The tech is SO far from being ready to replace human-driven vehicles as a door-to-door taxi service. Impressive safety records author mentions conveniently forget the fact that THERE IS STILL A SAFTEY DRIVER IN THE CAR. AV companies publish reports of how many 'disengagements' they experience per mile driven, but companies define 'disengagement' themselves--so its not clear that if the safety driver touches the brakes and or massages the steering to avoid an accident or make a maneuver smoother that this would ever be reported. Plus there is the problem that no one seems to understand that taxi / uber / lyft drivers do more than just drive! There are all kinds of customer service functions that need to be figured out, and in the near term these services are going to be slow, buggy, and annoying.

[waleonishmail]

lol

[PEAK AUTOS]

Very interesting article. How quickly do you think Waymo can have an ARS ready to go after placing an order with the car manufacturer verses Cruise? Say everything is going well for Waymo and they decide on expanding to Chicago How long will it take them to be operational in Chicago? Thanks

[Michael McGrath]

The orders were made a while ago, but the delivery schedule is dependent on the capacity of the Waymo factory to do the retrofitting of autonomous driving technology. I takes a while, maybe 6 months do complete the detailed maps for each new metropolitan area. But I don't think Chicago will be a priority for the next 3-5 years because of weather and difficult road systems.

[S_Archer]

The factories are not the limiter. the limiter is the fact that it doesn't work without a driver. this after years of work in Chandler.