There is a close parallel between the automotive and pay-TV industries when it comes to Android’s growing penetration. Automakers were facing challenges getting features to market quickly enough in the early era of internet connectivity, just as pay-TV operators did as online delivery became both an opportunity and a threat to their business.
Both faced new existential threats from some tech giants even if the major adversaries were different. For car makers, Google’s Waymo is a major threat, along with a new generation of electric car makers led by Tesla. There is also disruption to their traditional business model from Uber and others devaluing traditional private car ownership. In both these sectors, Google has developed an improved sector-specific version of the operating system – Android Automotive in the case of cars, launched in June 2017.
Before that, in 2015, came Android Auto, which was a direct equivalent of Apple CarPlay, focused on tethering smartphones – allowing customers to project apps such as maps onto the vehicle’s larger head unit display. These can be seen now as opening shots in a battle for control of the vehicle’s central console, and all associated infotainment or even control functions, which will play out in the run-up to autonomous driving, if that happens.
It was at that stage some car makers, especially Toyota, raised objections to this infiltration and accused Google of demanding too much data on parameters like vehicle speed, engine revs and throttle position. This too had echoes of pay TV, although operators there had the greater fear of Google sucking away their customers as well as their data.
Meanwhile though, Google decided that the smartphone pairing approach was really a dead-end and pursued full integration with Android Automotive, signing up Audi and Volvo as initial partners and customers. A bigger coup followed a year later in September 2018 when the so-called Alliance comprising Renault, Nissan and Mitsubishi signed a multiyear deal with Google to produce vehicles with in-built infotainment systems, based on Android Automotive, for launch in 2021.
Google is now going flat-out on Android Automotive, and curtailing its effort on Auto now, as sentiment turns against the tethering approach, even if the smartphone will still be a major component of the connected car ecosystem for connectivity, personalization, remote control and virtual car keys.
In fact, Google has unbundled some of the Android features from Automotive, in order to make it more flexible and more appealing to car makers seeking to differentiate their services, again with similarities to Android Operator Tier in pay-TV. Even Google Maps has been relegated to just an option, despite being still thought superior to rivals such as Apple Maps, and one of the major factors attracting car makers to Android in the first place.
Many consumers already accustomed to Google Maps on their smartphones, even if they are iPhones, have been disillusioned by the clunky navigation and sloppy graphics coming with their in-car mapping systems. They have naturally wondered why car makers don’t just bundle in Google Maps and that is what many have effectively done by incorporating Android Auto in the first instance, even if those not going for Android Automotive.
Android Automotive then does away with reliance on the smartphone for connectivity and instead uses an independent cellular connection, while providing a single framework for car manufacturers and developers to work with. In Volvo’s case, its Sensus infotainment system will remain intact, apart from enhancements, but be ported as a custom interface on the Android Automotive, just as can be done on smartphones and now Android-based set-top boxes.
This can then harness the connected infrastructure essential for connected cars, with the ability to store driving data and analyze it to tune drivetrain settings automatically. There is also support for communication between vehicles, as well as more services for real time updates on weather and traffic conditions for route optimization.
Of course, such capabilities are not unique to Android, and some of the major car makers are taking different routes, although in every case consolidating down from multiple proprietary operating systems to a common platform. Volkswagen for example appears only recently to have acknowledged the need to refine its operating software to a single OS and has rebuffed Android so far, apart from allowing Android Auto for pairing. It announced in August 2018 that it would develop a new operating system dubbed vw.OS as part of a $4bn program that includes a cloud computing platform, to connect vehicles and customers to services such as car sharing.
The idea was to create a system better suited than alternatives for running subsystems that will serve autonomous driving, while at the same time facing out to external services vehicle users will want to access. VW also argues that over-the-air software updates for cars could be executed more effectively if the operating system was designed in-house, rather than as at present depending on third-party software supplied by respective vendors providing various sensors. That is surely a given, but VW has not spelt out why it wants to go it alone at the OS level, when most of its rivals believe their resources are better devoted to differentiating features on top of something like Android Automotive.
Many rivals are building around Automotive Grade Linux (AGL), which is a collaborative cross-industry effort developing an open platform for the connected car, supported by Ford, Honda, Mazda, Nissan, Mercedes, Suzuki, and Toyota. The latter was first to commit for a specific vehicle when in May 2017 it announced AGL would be the platform for its next-generation infotainment system in the Toyota Camry launched this year. So it is boiling down to Android versus Linux on the car front with VW a notable outlier ploughing its own furrow. BlackBerry also lurks in the shadows.
Android has also been built around the Linux kernel and can run either as the primary OS as it does at Volvo and Audi, or as a subsidiary OS to Linux through some container or hypervisor. There are signs of this second option gaining traction, so this is not going to be a binary choice, and it looks as if while car makers picking Android Automotive will stick with that, those in the AGL camp may incorporate Android as well.