Deciphering Apple's Autonomous Vehicle Strategy
Michael E. McGrath
Originally published in Seeking Alpha on Jun. 13, 2018
Summary
For years there have been rumors about Apple's autonomous vehicle (AV) strategy, many leading to mistaken conclusions.
Apple recently received permits from California to increase its test AVs from 3 to 62, clearly showing it is serious about AVs.
Apple also recently agreed to customize Volkswagen T-6 vans into autonomous shuttles.
I believe that Apple will enter the market for autonomous ride services (ARS), although there are pros and cons to its success.
If successful (and there is still a lot to be proven), this could provide Apple with a significant new source of revenue.
In April 2017, Apple (NASDAQ:AAPL) was granted a permit from the California DMV to test three AVs on public roads, using Lexus RX450h SUVs. It just recently increased this to 62 authorized AVs in California, along with 83 permitted AV drivers. Apple now has the second largest AV test fleet in California behind GM Cruise with 104, although other companies such as Waymo are testing more vehicles in other states.
Recently, Apple was reported to be developing and preparing to test an autonomous shuttle van with its employees later in 2018 or 2019. It has a deal with Volkswagen (VLKAY) to customize its new T6 Transporter vans for these autonomous shuttles.
Over the last 18 months, I have researched autonomous vehicles, the technologies that enable them, the opportunities they will create, and the strategies expected competitors would follow. My earlier Seeking Alpha article - Google Tips Its Hand on Its Autonomous Vehicle Strategy, And It Will Be Big - introduced Waymo's (GOOG) (NASDAQ:GOOGL) AV strategy and the significant opportunity for autonomous ride services. Now in this article, I will try to decipher Apple's AV strategy and its opportunity.
My focus here is long term and strategic. It involves my expectations of the AV markets and Apple's strategy. It is not based on any insider information, but rather on my deep experience in the strategies of technology companies. It may be of interest to long-term investors in Apple who are wondering if there are any new significant revenue drivers on the horizon in the next 2-4 years.
Fragmented History of Apple’s AV Strategy
Apple keeps its product development strategies a secret, so one can only piece together leaks and rumors to describe its fragmented history, but even this provides some valuable insight. Apple reportedly began working aggressively on what it called project Titan by 2014, with what was rumored to be a large team of 1,000 or more. A couple of years later, Apple refocused project Titan away from designing and building its own complete autonomous vehicle, laying off several hundred people in the process.
Reports at the time claimed that Apple scuttled plans to build a self-driving car of its own, after a haphazard effort fell into disarray, to focus on the underlying technology. Frankly jumping to this type of conclusion fails to recognize how strategies for new markets evolve. Apple's refocused strategy made a lot of sense because at that point in time, insights were beginning to suggest that autonomous ride services (ARS), not retail sales of AVs, would be the most significant opportunity for a newcomer in the market.
In conjunction with this conclusion, there was also a recognition that the first generation of ARS vehicles would most likely be retrofitted versions of production cars. Waymo probably made a similar strategic shift around the same time.
In June of 2017, Apple CEO Tim Cook spoke publicly about Apple's work on autonomous driving software, confirming the company's work. "We're focusing on autonomous systems. It's a core technology that we view as very important. We sort of see it (referring to autonomous vehicles) as the mother of all AI projects. It's probably one of the most difficult AI projects actually to work on." “There is a major disruption looming there,” Cook later said on Bloomberg Television, citing self-driving technology, electric vehicles, and ride-hailing. "You've got kind of three vectors of change generally happening in the same time frame."
In the interview, Cook was hesitant to disclose whether Apple will ultimately manufacture its car. "We'll see where it takes us," Cook said. "We're not saying from a product point of view what we will do." The reference to ride-hailing indicated Apple's interest in autonomous ride services and the flexibility in manufacturing indicates keeping options open over the long term.
Toward the end of 2017, there were rumors that Apple was leasing a former Fiat Chrysler-owned proving grounds in Arizona to test autonomous vehicles. Also, there were indications that Apple was recruiting automotive test engineers and technicians from other proving grounds for that location. The property was leased to a company named Route 14 Investment Partners LLC. with no other identification or link to ownership.
Recent stories, particularly one published by the New York Times on May 23, criticized Apple for what they saw as a failure of Apple's autonomous vehicle strategy. The New York Times article was mistaken stating that: “Apple once had grand aspirations to build its own electric self-driving car and lead the next generation of transportation. Over time, the tech giant’s ambitions ran into reality. So, Apple curtailed its original vision, first by focusing on software for self-driving cars and then by working solely on an autonomous shuttle for its own use with employees. Now, the tech giant has settled for an auto partner that was not its first choice.”
A better understanding of autonomous vehicles and the new markets they will create provides a very different interpretation of Apple's AV strategy. Apple, like Google and other innovative companies, started looking at autonomous vehicle technologies and opportunities many years ago. Like most new technologies, as the technologies develop then the best ways to bring these technologies to market also change.
Apple’s Investment in Autonomous Vehicles
Apple does not disclose its R&D investment by product area, but based on rumors and some indicators, it appears to be investing significantly in autonomous vehicles and may have stepped up that investment recently. Apple has not disclosed when it started investing in AV development, but Apple was rumored in 2014 to have as many as 1,000 people working on project Titan, and then reduced that number by several hundred. So, its AV development project started well before 2014.
Over the last couple of years, Apple has been aggressively recruiting autonomous vehicle engineers and drivers. The typical total fully-loaded cost for an engineer is approximately $200,000 per year, so 750 engineers cost roughly $150 million annually. Other expenses such as labs, equipment, sensors development and testing, pilot test vehicles, and potentially a test track, can quickly add up to $50 million or more.
This would put Apple at investing close to $200 million annually in autonomous vehicles, which is approximately $1 billion over the last five years. Testing autonomous vehicles can get very expensive with the capital investment in the retrofitted vehicles and drivers. With two more years of pilot testing, Apple could potentially invest cumulatively as much as $2 billion developing autonomous vehicles by the time I expect the market to emerge in 2021. And it could be investing much more if it is designing a specialized computing platform for AVs.
Another data point on this potential investment is the acceleration of Apple's investment in R&D. Its R&D investment has jumped significantly from $4.5 billion (3% of revenue) in 2013 to $11.6 billion (5% of revenue) in 2017. While this by no means is meant to imply that most of the increase in R&D investment is all in autonomous vehicle R&D, it does suggest that Apple is investing in much more in R&D than the sustaining engineering of its existing products such as the iPhone, and there is plenty of room for a $200 million or even greater annual AV investment in the $7 billion increase.
This investment does not include Apple’s $1 billion investment in May 2016 in Chinese ride-hailing service Didi Chuxing, which dominates the ridesharing market in China. Apple later received a seat on the Didi Chuxing board of directors. The ultimate benefit of this strategic investment is still to be seen, but it may provide Apple with some options of partnering in the Chinese ARS market.
I believe that Apple intends to compete in autonomous vehicles (AVs). The various AV markets will be enormous, more than $1 trillion in the US alone by 2030. As one of the largest high-technology companies in the world, Apple cannot afford to sit on the sidelines and let others succeed. AVs are probably one of few, maybe the only, new market in the next decade that will enable a company of Apple's size to grow significantly.
The real question then is what AV strategy we can expect from Apple.
What Apple’s AV Strategy Is NOT
In deciphering Apple’s potential AV strategy, we can first start by looking at what its strategy is NOT. Apple will NOT design, manufacture, and sell cars by retail to the public. This market is just too difficult, risky, competitive, and expected to shrink. Apple will NOT acquire Tesla (TSLA) – sorry to disappoint all those investors hoping for that. Tesla, Ford (F), GM (GM), BMW (OTCPK:BMWYY), Mercedes (OTCPK:DDAIF), and others will fight it out in this market. I also believe that the AV retail market will take much longer to emerge than other AV markets. Apple does not want to get involved in this market.
Apple will NOT simply license its AV technology. While it has invested billions in developing AV technology, so have others. It’s unlikely that these others will simply abandon their AV investments and license Apple’s. Plus, it is not a very lucrative opportunity. Even if Apple could license its software for $2,000 per vehicle (not likely), and could license this to 500,000 to a million cars per year (which would be a high estimate), this is only potential revenue of $1 to $2 billion revenue, at best.
The entertainment capabilities inside the vehicle will be important, but the opportunity for Apple to license this to auto manufacturers is also a relatively small opportunity. Most of the ARS companies and AV retail car manufacturers will want to control their own in-vehicle entertainment and get the revenue stream from this.
What Apple’s AV Strategy Probably Is
The most attractive new AV market for Apple is autonomous ride services (ARS). It is a brand-new market with no entrenched competition. It will be the first significant new market for autonomous vehicles because the vehicles simply need to be sufficiently autonomous to succeed. Technology companies such as Google and Apple have some critical advantages entering this market.
It is a new market, so there are not any well-entrenched competitors, other than Uber (UBER) and Lyft (LYFT) which are providing similar, but much more expensive, ridesharing services. A new market enables the early entrants to define the market to some extent and not have to compete on the terms of incumbent competitors. The ARS market doesn’t require the supply chain and retail channels of distribution that create barriers to entry into the retail AV sales.
The ARS market will require an autonomous vehicle platform of technologies – specifically the sensors, computing power and most importantly the software for autonomous driving. Technology companies have a significant advantage in creating this platform. The vehicle itself is also part of the platform, but this part is a "dumb" vehicle that can be a commodity or eventually custom-designed vehicle specifically for ARS.
Rides in the ARS market will be requested using an app on iPhones or other smartphones. Apple and Google control the operating systems for smartphones and can add their ARS apps directly to the smartphone with an operating system update. Also, Apple can automatically link its new ARS app to almost a billion Apple Pay and iTunes accounts.
The Autonomous Ride Services (ARS) Market Opportunity
In my recent Seeking Alpha article - Google Tips Its Hand on Its Autonomous Vehicle Strategy, And It Will Be Big - I described my thoughts on the emerging market for autonomous ride services (ARS). There were many questions about the potential size of this market, so I will try to provide more detail here.
ARS is ridesharing without a driver. The economic advantages of eliminating the cost of a driver and achieving high utilization of ARS vehicles will reduce the cost significantly. I estimate that the price of a typical ARS trip will be less than half of a typical Uber trip today, and it might be even much lower than that. Even at this much lower price, ARS will be very profitable compared to the regular loses in ridesharing. ARS will be even more convenient than ridesharing because rides can be scheduled in advance, and ARS vehicles will provide a more comfortable interior.
ARS will be the first market for autonomous vehicles (AVs). People will be more comfortable trying out a trip on an autonomous vehicle than taking the risk of purchasing one. ARS will be initially provided in selected municipal areas, most likely in the South and areas with favorable weather conditions. ARS vehicles only need to be sufficiently autonomous, by which I mean that they just need to be able to travel on specific routes, not every back road or alley. Geo-fencing will define the suitable roads for ARS vehicles.
I expect that ARS will be fleet-based. So initially, 500-1,000 ARS vehicles will be placed into a typical municipal area by an ARS company with potentially 2-3 ARS companies competing in each market. Because they are fleet-based, the ARS vehicles can be more easily serviced, maintained, and cleaned. An ARS will use specifically defined routes within each metropolitan area that its vehicles can maneuver those routes with proven reliability. When a passenger requests a ride from one location to another, the ARS app will validate that it is a route it can manage, and if not, then it will not offer that ride.
For many people, ARS will reduce the cost of transportation. AAA estimates that it costs approximately $8,500 to own a car, not including the cost of fuel. With the cost of fuel, this translates to about $850 per month for a car, not including the cost of parking, which is considerable in many cities. Many people will find it less expensive to use ARS for their travel needs, especially families who may own two cars. There have already been studies that using Uber is less costly than owning a car in several American cities. With a much lower cost per trip, the number of cities where ARS is less expensive than owning a car will increase significantly.
If this happens, then car ownership and new car sales will decline, but this is a topic for a later article.
ARS at its lower cost may also start to displace public transportation. If you look at a range of examples, the cost of an ARS trip could be comparable to public transportation, and it eliminates the need to walk to and from the bus or train stop by going directly from where you are to where you want to be. I also expect that some businesses, particularly restaurants, will offer complimentary rides to their customers. In many cases, it will be less expensive than valet parking and will attract more customers and potentially increased alcohol sales.
ARS will be such a potentially large and profitable market that I expect that several large companies, including Apple, will enter this market aggressively. The nature of this market will also motivate competitors to open new municipal markets quickly. I compare it to a land rush. If Waymo and Uber each put 1,000 vehicle fleets into a specific municipal market, then Apple may bypass this market and go into another. Once ARS is accepted as a viable and attractive service, aggressive competition will drive rapid growth.
I estimate the size of the ARS market in the United States to be approximately $150 billion by 2025. I get there with my assumptions that each ARS vehicle will make 50 trips per day, which is the same assumption that Waymo subsequently made, and that each trip will average $8.75 with vehicle utilization of 350 days per year. This equates to approximately $150,000 of revenue per vehicle annually. With 1 million ARS vehicles operational by the end of 2025, total ARS revenue in the US would be $150 billion. It would still only account for approximately 3% to 3.5% of total miles traveled.
Another way of looking at this is through the projected growth of ridesharing. The number or ridesharing rides in the United States last year, based on reports for Uber and Lyft, was approximately 2.5 billion. Ridesharing grew at a fantastic pace, averaging more than 75% from 2016. Depending on continued growth assumptions, the ridesharing market in the US by 2025 could range from 43 billion trips (at 50% growth) to 80 billion trips (at 70% growth). My assumptions for the ARS market would equate to 17.5 billion ARS trips in 2025. This would be reasonable cannibalization of the ridesharing market even at the 50% growth rate, and this doesn't include market expansion due to lower-priced trips and increased convenience.
Why Apple's Previous Negotiations Were Not A Failure
The key to success in the ARS market, as it is in many technology markets, is a platform strategy where you provide or control the critical aspects of the product or offering platform. In the ARS market, ride request and dispatch sit at the top of the offering platform. Whoever controls this, can provide or subcontract the actual autonomous vehicles.
The critical value-added, or what I refer to as the defining technology, in the AV platform is the autonomous driving capabilities, specifically the sensor package, computer processing, and the extensive software to direct the vehicle. Except for unique interior designs, the actual “dumb vehicle” is less important and can be replaced over time. It essentially becomes a fleet purchase or contract manufacturing.
Companies who want to compete in this market realize the importance of this platform strategy, and they don’t want to be put in a subordinated position. They want to control the platform. The New York Times characterization of previous Apple negotiations with BMW and Mercedes as a failure misses the strategic issues.
These companies don't want to be put into a subordinated position to Apple with Apple controlling the customer, in this case, the ARS passenger. In fact, they intend to compete with Apple. Apple wants to control the critical aspects of the ARS platform. So, one company needs to give into the long-term success of the other, and neither company wanted to do that.
Why the VW T-6 Transporter Makes Sense for Apple
The Volkswagen T-6 Transporter is an interesting strategic choice for Apple. It is a very popular vehicle in Europe, although not sold in the United States. It is also highly configurable with different seating arrangements, but it is also easily customized to individual preferences.
According to the New York Times article, Apple and Volkswagen plan to remake the T-6 into a custom autonomous vehicle, potentially even converting it into an electric vehicle. The basics of the T-6 will remain - but seating, the interior, and even the dashboard may change. This is in addition to adding the sensors and computers required for autonomous driving.
Apple will initially use the vehicle as a pilot to shuttle employees between its two main campuses with a backup driver and technician tracking the vehicle’s performance. Once this autonomous vehicle is proven, Apple could include a customized version of this autonomous vehicle into its autonomous ride services fleet in addition to the Lexus RX450h SUVs or other autonomous vehicles.
It is also possible that Apple could provide this shuttle for autonomous shuttle services in some partnership form. I think autonomous shuttles also will be an early AV market because they only need to be sufficiently autonomous to go from one predetermined location to another along the same routes all the time.
The Timing For ARS Market
I expect that the ARS market will be created in stages. We are currently in what I refer to as Stage 0 (2016-2020), Development and Testing, and this stage will continue until 2020. During this stage, we will increasingly see more ARS pilot programs, using autonomous vehicles without drivers.
Waymo has indicated that it plans to test ARS without a driver starting later this year. GM Cruise intends to test ARS in 2019. Apple's intention of testing its shuttle service in either this year or early next year, as well as its rapidly increasing number of ARS vehicles being tested in California, are consistent with being competitive in this Development and Testing Stage.
I see Stage 1, the Launch of ARS in the U.S., as starting in 2021. This is the big transition point when ARS will become an accepted form of transportation in many municipal areas. By the end of 2021, I estimate that 100,000 ARS vehicles will be in service, generating approximately $15 billion in revenue.
This estimate may prove to be low since Waymo has already indicated it will be testing 20,000 I-Pace SUVs and 62,000 Chrysler Pacifica Hybrid minivans by 2020. From 2021, I expect that the number of ARS vehicles will double or more annually, reaching 1 million vehicles and a $750 billion market by 2025. At this point, it will still only represent about 3% to 3.5% of the total miles traveled in the United States.
During Stage 1 (2021-2025), I expect that most of the ARS vehicles will be customized versions of car models currently in volume production by car manufacturers. The level of customization may vary. Many will still have steering wheels, while some may have significantly customized interiors designed for passenger comfort.
These customized vehicles will mostly be "dumb cars" without any ADAS functionality that are purchased in volume with the autonomous driving capabilities retrofitted by the technology company onto the vehicle. In particular, I don't expect Waymo, Apple, Uber, and Lyft to get into car manufacturing. The auto companies entering the ARS market may have somewhat different strategies.
During Stage 2 (2026-2030), what I refer to as Broad AV Acceptance and Disruption, the ARS market will accelerate. This is also the Stage where I expect fully-automated vehicles to be sold in volume to retail customers. Focusing on the ARS market in Stage 2, most ARS vehicles will be custom designed. They will look more like living rooms or offices than cars. The technology inside for passengers will be impressive. The ARS companies will begin to design these custom vehicles during Stage 1 and contract with manufacturers, either auto companies or contract manufacturers like Magna (MGA), to build them in volume.
I estimate that the ARS market will reach 5 million vehicles by 2030 and generate approximately $750 billion in revenue in the United States, but it will still only capture about 15% of the total miles driven. Globally, ARS will be a huge market with more than $1.5 trillion in revenue in 2030, causing enormous disruptions to other industries.
Can Apple Be Successful In The ARS Market?
While I have confidence in the size of the ARS market and confidence that Apple will enter this market, it is premature to allocate market share and determine the winners and losers in this market.
Waymo has some early-mover advantages that could translate into early market share advantage. As I mentioned, Waymo expects to begin testing its ARS this year and has commitments to buy as many as 80,000 AVs over the next few years. I don't see the ARS market emerging for another 2-3 years, so there is still time for all competitors. Also, I envision the ARS market to be deployed regionally by metropolitan area, and it will take 5-10 years to fully deploy throughout the US, giving time for later entrants to still capture meaningful market share.
There are also skeptics that see Apple's shortcomings in artificial intelligence as carrying over into the software that controls autonomous vehicles. Apple could fall so far behind Waymo that it might never be comfortable to enter the ARS market or be so far behind that it never captures meaningful market share.
There will be other competitors too in the ARS market. GM's recent move to carve out its Cruise business targeting ARS with SoftBank (OTCPK:SFTBY) investing $2.25 billion and GM investing $1.1 billion demonstrates its intention of being a serious ARS competitor. Uber has serious intentions in the ARS market, as it must because ARS will almost wholly cannibalize ridesharing, but it has been stumbling recently. I also expect Ford and Lyft to be serious competitors, and there are other possible competitors too.
The ARS market will be very large, and I expect intense competition. A meaningful market share of this $150 billion market by 2025 could provide a significant revenue for the companies that succeed. A 20% market share would generate $30 billion in revenue, and a 33% share could generate $50 billion in revenue. By 2030, even a 20% market share could translate into $150 billion in revenue.
As I mentioned in my previous article, this industry will be capital intensive. Google and Apple will have some advantages with their large cash balances.
Unlike most of its products, Apple will need to disclose its ARS strategy in advance of a product launch. This is because of regulatory approvals, local metropolitan area approvals, and the size of the vehicle investments. I expect that Apple will disclose its ARS strategy in the next 1-2 years, so it can be a viable competitor when the ARS market emerges in 2021. At that point, investment analysts will begin to build ARS revenue estimates into their projections. Prior to then, it will be necessary to decipher Apple's autonomous vehicle strategy to get a jump on that.