Silicon Valley – the New Motor City?
As you may have noticed, my last few blogs either directly or indirectly, have been nipping at the issues surrounding autonomous vehicles (AV) and the spatial data that might be needed to operate them. Today’s post offers a look at how Google, Apple, Uber and others might choose to compete in this market.
While the current automobile manufacturers and their suppliers obviously will attempt to compete in the AV market, it remains an open question as to whether or not these industry stalwarts will be able to effectively transition to a future AV market, in which vehicles being driverless,” and manufacturers being dealer-less may not be the most significant changes. The complex infrastructure that supports the motor vehicle industry (e.g. suppliers, dealers, repair shops, parking facilities, aftermarket services etc.,) will likely experience disruptive change as AVs emerge. In turn, this transition will generate enormous money-making opportunities related to the new and potentially unique infrastructure requirements required to support the AV market. (See this article from Forbes for some interesting insights. )
The implications of AV market development are incredibly complex and most of us have not thought through the changes that will accompany this market. Will cars that cannot crash need to be insured? How will local municipalities replace the income generated by motor vehicle violations when traffic fines disappear due to vehicles being programmed not to violate local transportation and parking ordinances? Will it be more economical to run fleets of AVs around the clock (as delivery vehicles in off hours), rather than park them overnight? Will houses need garages for vehicles? How will the transition be managed when SUVs and AVs share the highway before AVs replace old fashioned drive-it-yourself vehicles? Who is going to employ those taxi, truck, bus and other commercial vehicle drivers made redundant when AVs become commonplace? Obviously the questions are endless – as are the opportunities.
It is my position that what happens to the existing market for vehicles, its infrastructure in general, and current OEMs in particular, will depend, to a considerable degree, on how Google, Apple and Uber attempt to monetize and leverage the market for autonomous vehicles. I realize that some of you think that this scenario is implausible. Others might respond that this is the same mindset held by mobile phone manufacturers (e.g. Nokia and Motorola) when contemplating the news of iOS and Android-powered mobile phones. Note that disruptive innovation rarely comes from existing players in a market.
In this blog I do not intend to include a long background discussion on AV, as any search engine will result in tons of information on the technology. Instead, the paragraphs that follow outline how I see the future AV market unfolding mainly for the three key players that interest me today. Let’s start with market timing and then move on to Google, Apple and their approach to the AV marketplace. Comments on Uber’s strategy will be based on the potential strategies of Apple and Google.
While there is an amazing amount of ongoing work aimed at producing saleable AVs, it is likely that mass-produced autonomous vehicles are over a decade away. Before then, we will see semi-autonomous vehicles that require varying degrees of driver intervention. AV prototypes will become abundant during the next five years, but the market will remain miniscule over the next decade. During this market development period the existing manufacturers will attempt to show their ability to innovate, as well as their influence while lobbying for legislation that might tilt the table in their favor over companies threatening it from the software/technology worlds.
The possible strategies of Google and Apple
Google makes the Android OS, billed as the world’s most popular mobile OS, available to everyone, but produces and markets their own version of an Android phone to show mobile companies and software developers how the system should work. The original Android, aimed at creating an Open Source OS for mobile phones, was acquired and managed by Google to further its corporate goals, while sparking a communications revolution. In respect to those “corporate goals,” Android is advertised https://www.android.com/intl/en_us/ as having, “The best of Google built in,” which means that “Android works perfectly with your favorite apps like Google Maps, Calendar, Gmail and YouTube.” What that really means is that Google has you in the clutches of their massive and highly profitable advertising business.
Apple, on the other hand, developed its iPhone operating system (iOS) with the intent of delivering it exclusively for use in a handset developed by Apple and manufactured to its specifications. Apple’s amazing financial success is based on the fact that the company’s products are designed to provide a comprehensive user experience for its customers. In turn, Apple’s customers rely on the company’s integrated approach to product development, equipment manufacturing, support and device education, as it provides them an “upscale” product experience.
The approaches of Google and Apple to the mobile market potentially speak to two potentially different strategies for competing in the AV market. It seems to me that the goals of Apple and Google in respect to the AV market should be the same – they should not care about the car, they should care about controlling what goes on in it when cars become autonomous. It is at this point that, for the operator, the vehicle becomes a floating- office/living room/den/bar/restaurant etc. Relieved of the “duty to drive” people will want to use the car cabin for recreation/communication/lifestyle experiences or to conduct business. I am not sure that either Apple or Google needs to own the car to own the cabin, but Apple may well want to own the “car” and the entire vehicle experience.
Google may conceptualize the vehicle cabin as a “local search petri dish” – abounding with germs – each of which is a new thread opening a unique path to sell local advertising. Apple may see the car as another device whose sales will dramatically increase their income and include a cabin-based market for services that a captive audience will be willing to buy/lease/rent.
Google and its Johnny Cab
It is my opinion that Google is looking to kick-start their concept of how the AV World should work through the development of what will become a low cost, fleet-oriented solution to navigation that encompasses the concept of cars on demand and no need for individual ownership of the vehicle. While these are laudable goals, they may be so utopian in scope as to preclude success. The problems with implementing such fleets are enormous, but, then again, those who think of ways to do so will make fortunes.
Of course, Google’s goals are flexible and the company will likely produce a purpose-built automotive oriented OS running its Johnny Cab AVs to show other manufacturers how a Google-based system could benefit the development of AVs. Specifications for its Johnny Cab OS, including the details of the hardware, sensors and data (e.g. a map database, traffic database, vehicle restriction database) required to make the vehicle truly autonomous will also be available. However, it will allow its licensees adopt the OS to fit their own vision of AV production and style, just as it does today with Android in the mobile phone market. Google will, however, control and tightly license use of the databases key to operating the system. In this strategy, Google may become the best friend of existing automobile manufacturers and their suppliers, although providing the AV OS and support infrastructure will put Google, not the vehicle manufacturers, in the catbird seat. See Google’s pre-marketing for wireless cars here.
It is my opinion that Google already possesses the majority of tools, know-how, and databases required to create fully functioning AVs. Tuning their OS and optimizing the vehicle to meet their view of the future will take some time, but certainly they are the lead horse in the AV race. Their key concept of ride availability based on vehicle fleets rather than individual ownership may be further in the future, but the notion does solve a number of difficult mobility-related problems that society faces today.
Google’s strategic desire in their AV development effort is to ensure the production by others of a large population of AVs powered by Google as a method of extending the sphere of influence of its advertising business. I doubt that Google is interested in becoming a manufacturer of automobiles, just as it has not shown much interest in being a manufacturer of mobile phones, tablets, or computers. Apple, on the other hand, well, Apple will likely take an entirely different approach.
Apple and its iCar.
In some ways, Apple appears to be more realistic about exploiting the market, sensing that people want better designed products – an iCar OS for instance. Apple helped transformed the world of telephony into a world of social contact, software apps and declining voice communications. The company has similar dreams for transforming the AV world into a social experience that transcends the friction of distance by focusing each vehicle trip on you and your wants during the journey, not about the vehicle, or how to get it to your destination, or how to avoid other maniac drivers.
Apple’s presumed approach, however, requires more than just an OS. The iCar must be a complete user experience that Apple can control from design though production. Apple will likely want to develop, market and support its own branded vehicle and is currently reported to be executing this concept at a skunkworks in Silicon Valley .
In essence, Apple may become the real threat to the current automotive manufacturers and their suppliers. Fortunately for them, Apple’s penchant for high-end, high-margin products will make it likely that Google or a Google-Like Company (GLC) will partner with the existing industry bringing the IOS vs. Android wars to the segment of the new marketplace that does not require “luxury” vehicles.
While hardware for AVs is not a crucial question (for an overview of the required hardware – see this from Wired) I have some doubts that Apple is up to the “whole car” challenge. Apple seems to be shy about working with anything that is mechanical in nature. Think of the evolution of all of Apple’s devices and you will notice that over time they have become smaller and less mechanical (dials replaced by touch screens, etc.). While I think Apple understands the world of “mechanical things”, I believe that the Company considers them the weak link in all use-scenarios. I wonder if Apple would be familiar with managing the life cycle of a mechanical automobile and dealing with the problems that such systems might present to their customers and to Apple as a product/service provider. Today, if you are having trouble with your still under warranty iPad, you go to an Apple store and once they confirm the fault, they simply swap out your unit. This type of service might be more difficult in the future AV market, but a modular car and component system could be maintained by remote diagnostics and serviced by “hot-swapping” equipment centers. Time will tell.
Note, that Apple’s map database might need some serious augmentation in order to support the company’s AVs. Today its database lacks the “map” details needed for autonomous navigation and also lack the roadway/roadside imagery and lidar data that could provide a significant benefit in the AV market, items that Google has been industriously collecting for some time. On the other hand, everyone involved in the potential AV market needs to ask themselves whether “map” database systems architected and designed for high-volume, Internet-based map serving will meet the needs of managing the tasks required to simultaneously, safely navigate million of AVs. I suspect there will be significant performance difficulties when using these map databases in AV applications – but that is not a topic to be covered here – at least not today.
And Now – Uber
Today Uber does not need to manufacturer or otherwise own a communications device or OS to support its business, since the phone and related apps are simply regarded as connecting the user to the Uber infrastructure. However, in the future, their infrastructure will be significantly transformed by AVs, their adoption rate and the nature of their ownership (personal or commercial). If my guess about the strategies of Apple and Google are correct, then Uber may have a difficult time competing, at least in some geographic markets, unless they reconfigure how their business operates.
Uber connects rider with vehicles and drivers. At its heart, Uber depends on driver/vehicle availability and matching these with the demand for trips generated by passengers. It may be that this scenario will become more difficult to manage if the scenarios for Apple and Google actually occur as outlined above. Owners of an iOS AV would, I think, be unlikely to make them available for use by Uber. Fleets of Google’s Johnny Cabs would already be programmed for maximum availability and, likely, not be available for use by Uber. While this issue could be considered a matter of logistics, Uber could take the uncertainty out of the equation by producing and owning a dedicated fleet of AVs. In this scenario Uber would continue to support its present business, but through its own fleet of AVs. The company could, also, expand its offering on a contract basis to provide individuals with transportation services that eliminate their need to own a car at all. Finally, Uber could add to its street visibility and market strength, by designing and fielding a fleet of custom AVs, in much the same way that the UPS brown trucks are unique and identifiable anywhere in the world.
It is likely that the notions described above are what has caused Uber to bid for HERE, as well as to partner with Carnegie Mellon on the Uber Advanced Technologies Center. From my perspective Uber could save a lot of money by being a licensee of Google’s forthcoming AV products, but, apparently, the company’s strategic interests in new markets makes them reluctant to partner with a likely competitor. At this point, however, Uber’s investment choices do not convince me that they are on track to play a winning hand in this game of chance.
I think the key concept behind Google and Apple’s participation in the development of autonomous vehicles involves each company capitalizing on the freedoms that will result when automobile users are finally relieved of the duty to drive. Driving is a demanding, tiring task that provides few rewards other than to transform space by spanning the distance between an origin and destination faster than by possible using other personal transportation devices. Those finding exceptional ways to fill the time that people will no longer spend driving vehicles will find a number of incredible new markets. Aftermarket modifications will become even larger in the world of AVs. Privacy? Don’t even ask.
For Uber the future of the cabin is largely a continuation the past. Its current users are those who have already chosen to be relieved of the duty of driving. What will change for Uber is the potential need to own a fleet of vehicles rather than contracting this availability from independent owner operators in order to sustain their business. The choice of whether to develop their own AV or not will profoundly influence the future development of this dynamic company. Total cost of ownership (TCO) will become a fundamental part of their business model, and one that they may wish they had avoided. They will find that paying drivers to use their own cars provided a much better return.
And one final note. How are all these quasi-sentient AVs going to find their destinations when they are not residential addresses? No company that I know of has either comprehensive or accurate business listings data, a theme that we have long hammered on in this blog. Add in the growing mold of link-rot and you have a toxic problem for the AV industry. Let’s face it, the AV will not know where the driver actually wants to go, it just knows which item on the list that was presented to the user was selected as the destination. In fact, the AV rider may not know this is the wrong business name/address pair until they arrive. What fun – a vehicle that finds destinations using a spatially indexed random numbers table. So what else is new?