The auto industry is at cross roads. From having provided people unprecedented freedom of movement, it now struggles with pollution, accidents and urban congestion. At the same time there is a technology revolution on-going, where software and online feature upgrades are entering what used to be a piece of static electro-mechanical hardware. The auto industry is in a shot-gun wedding with IT technology that will bring about fundamental changes in the industry dynamics. Beyond the auto industry, this may even change our society at large. In a recent Ted Talk, Uber’s co-founder and CEO Travis Kalanick discusses how changing the way people use cars with IT can change the society for the better, by having fewer cars out in the road for the same amount of people in transit. In this blog, we take a look at how IT is disrupting the automobile industry, and what the implications are for the future.
Mobile apps have become a big part of everyday driving. GPS and navigation apps like Waze or Kimgisa (current Kakao Navi) show real-time traffic data, suggesting the best route to one’s destination. These apps crowd source traffic flow information, making routing dynamic and real-time optimized. Ride-sharing apps like Uber or Lyft are successfully shifting car-using habits through their “shared economy” business model. People increasingly consider whether driving one’s own car or taking Uber or Lyft is the better alternative. Incumbent automakers want a piece of the action, with the likes of GM and Toyota investing heavily in these companies. Beyond ride-share, car renting apps like ZipCar or Car2Go, offer another alternative to owning a car. By combining car rentals with the functionality of a mobile app (providing a mobile key to vehicles), these companies allow users to flexibly rent a car wherever and whenever needed. These business models are also spreading rapidly. Many Western apps are quickly copied into localized Asian versions, like Korean Kimgisa taking after Waze, or Chinese DiDi following Uber.
While mobile apps are changing how we travel by car, there is more happening within the cars themselves. With software managing controls and engines, and touchscreens entering the dashboard, car functionality is increasingly driven by IT. Operating Systems (OS) are becoming critical for user interaction, and major software giants, like Apple and Google, are competing fiercely to enter this area. Both of the major automobile OS, Apple’s Carplay and Google’s Android Auto, promise to provide a seamless infotainment experience from phone to car. They also promise to connect to the Internet of Things (IoT), enabling smarter sensing, access, and control of the vehicle. Interestingly, the most dominant player is Blackberry, of past mobile fame. QNX is a Blackberry subsidiary which offers a higher-level OS that can run either Apple’s Carplay or Google’s Android Auto on top of it as middleware, and it owns more than 50% of automobile OS market share.
Software is also connecting cars in new ways; Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) communication is entering. Once cars are able to communicate their location, heading, speed, and navigation route to neighboring cars, a large number of accidents can be avoided. In fact, the National Highway Traffic Safety Administration (NHTSA) predicted that V2V technology could prevent more than half a million accidents and more than a thousand fatalities in the United States per year. The same goes for V2I technologies, which has the potential of reducing congestion and make overall driving safer.
IT is also taking over the role of the driver, as self-driving cars (autonomous vehicles) are entering the market. Already today multiple technologies exist to assist drivers and take over some of the burden of driving. Building on aerospace, aviation and military technologies, self-driving technology is entering with force. Cheaper sensors enable more widespread commercialization such technology. Figure 1 shows different sensors needed. Note how much the car is part of the Internet of Things – thoroughly connected and aware of its surroundings.
Making self-driving cars work is more complex than implementing autopilot for planes or spacecraft, given moving obstacles like pedestrians and other cars, and also the complexity of traffic laws. The guidance algorithm becomes the soul of self-driving cars while the sensors are the eyes and ears. Advanced computing powers and artificial intelligence machine learning technologies are key for the car to understand, interpret, and make split judgments on the road, similar to what a driver would do. For machine learning to work, a massive amount of data from numerous sensors must be crunched to make the right decision in complex circumstances. Tesla Motors was able to get around this by pre-installing sensors in its cars and collecting data for ten months before releasing a software update that unleashed its “autopilot” capabilities. This shows how automakers today must be IT development savvy, thinking ahead and leveraging their installed base and “beta users”, just like software companies do.
No surprise then that the leading autonomous driving technology companies like Google or NVidia are from the software industry. Even Tesla Motors has a background in software, with its founder and CEO Elon Musk having founded two software companies before starting Tesla. However that does not mean that traditional automakers are idling behind. Mobileye is an IT company that develops vision-based driver assistance software and is used by many automobile manufacturers like General Motors, Volkswagen, and Renault-Nissan. The contestants for dominating self-driving cars include now radically different software-first cars like Google’s, but traditional cars with self-driving functionalities are also in the race, with many powered by outsourced technology.
With these novel software technologies taking over the auto industry, the way to “consume” cars is evolving. The concept of ownership will change, with car-sharing and fit-for-purpose cars. Just like consumers are abandoning cable TV and jumping to paid media content services like Netflix, a de-bundling of personal transportation services is underway, providing transportation-as-a-service rather than selling an entire car. The current model of personal car ownership is inefficient, as a perfectly functioning transportation machine spends 95% of its lifetime parked. Shared ownership and transportation-as-a-service help to reduce congestion and parking problems. People can select and use different types of cars based on their needs; a one-person vehicle for personal transport, a larger vehicle for family outings, luxury cars for special occasions, and trucks or SUVs when moving a lot of baggage. Traditional car manufacturers predicted this years ago, but did not realize that the way to the future was to be very different from just selling more cars to people.
The car lifecycle is changing too. Instead of selling an old car and buying a new one, software upgrades will be available for updating the functionality of a car (and extending its life), similar to how mobile phone operating systems (and even internet connected home electronics) are upgraded. Tesla is already utilizing this technique in lowering the entry barrier for purchasing electric cars – they sell shorter range cars at a lower price range but with hardware for extended range built in. Users can buy upgrades to increase range by unlocking the full hardware capability. A similar approach is used for self-driving functionalities. The user can buy the car at a lower price and functionality, and then upgrade later by purchasing a software key. The same applies to new features that come available as software is developed further.
The traditional automobile industry is notorious for having had a high entry barrier. Putting together thousands of components to manufacture a machine on wheels, and distributing and servicing the cars globally, are no easy tasks. Outsourced manufacturing and software are changing this, allowing new players enter. The supply chain is also changing; companies like Rightware, a dashboard software producer, have been able to find their spot in the automobile value chain serving the big automakers. They were able to do so by focusing their offerings on advanced niche areas where incumbents simply failed. Expect the industry to continue changing towards something like the tech industry (with myriads of startups and fast innovation cycles).
Software is also changing the ways of designing and producing cars. Designing automobiles is traditionally a multi-year process that requires huge amounts of resources and man hours. Recently a software expert was able to design a fully functional car in only three months that can go up to 100 mpg, utilizing an agile software development technique called “scrum”. This clearly shows that the auto industry is capable of shortening time-to-market in introducing revolutionary new designs with software development-like methodologies.
With autonomous vehicle technology rushing forward in today’s remarkable speed, we are eventually going to encounter the transition where we completely entrust software to take the wheel. The benefits that come from this greatly outweigh the alternative, as it will effectively reduce accidents and fatalities caused by human error. Consequentially, the dynamics of how we view and use cars will alter to fit the new norm of personal transportation. Cars with manual driving functions (like a steering wheel and brakes) will eventually become a niche market for recreational purposes and the act of driving will become more of a sport than a necessary utilitarian skill. This is similar to how manual transmission cars still exist in the market even though automatic transmission technology is considered the more practical choice.
There is big concern that professional drivers, like taxi or truck drivers, will be replaced with self-driving software. However the effect will be diluted due to the change being gradual. Figure 2 illustrates the different levels of self-driving technology, and how changing into autonomous driving is a step-by-step process. Drivers that fully control their cars today will transition into drivers that control certain functions of their vehicle, leaving the rest to automation. Further in the future, the same driver will become a vehicle supervisor, who oversees the driving to make sure the system does not malfunction and interferes manually only when needed. In addition, especially in the present day trucking industry, there is a huge shortage of drivers. Autonomous vehicle technology will actually enable more drivers to apply for the jobby lowering the qualifications and physical requirements.
Supplementary industries surrounding the auto industry will observe a shift as well. Auto insurances will have to change as auto accidents due to human error will decrease significantly once self-driving cars take over. The responsibility of driving safely will shift from the driver to the car and car manufacturer and accidents related to software malfunction will have to be dealt with. The recent fatal accident involving Tesla’s Model S on autopilot demonstrates the change in liability of these situations. Also in-transit entertainment and media consumption will expand drastically, as drivers will become passengers, freed up to actively consume more entertainment during their transit.
There are still practical obstacles to overcome before we can fully embrace the automotive software revolution. From being a mechanically centered business, the industry players must adapt to a more software-centric approach, both for the product itself, as well as the way of working and developing products.
Cyber security for cars will be increasingly important, and can be expected to grow dramatically. As part of the Internet of Things, keeping cars safe from possible attacks will become critical. Autonomous driving technology enables software to control the vehicle. Losing control due to a cyber-attack could be disastrous. The famous hijacking of a Jeep Cherokee’s software to the point that the hacker gained complete control was a chilling demonstration of how potentially terrifying cyber-attacks on automobiles can be. To ensure security of vehicles, cyber security for automobile software is going to grow fast, to match the development in vehicle software and autonomous technology.
To enable fully automated cars, policy makers will have to devise a smart plan to manage the transition from human-driving to self-driving cars. Partially autonomous vehicles (levels 1 through 4 on Figure 2) may pose potential threats to safety when unregulated, as the line between driver responsibility and vehicle responsibility in driving is blurred. Accidents are prone to happen when the driver decides to give more control to the vehicle than intended. A recent example is the many accidents that followed Tesla’s autopilot release, due to drivers trusting the system so much that they neglected their role in steering and driving the car. Following the recent fatal accident on a Model S, Tesla has announced that drivers are at all times responsible for keeping an eye on the road and hands on the steering wheel since it is not intended to be a fully automated driving function (level 2 out of 5). Also allowing fully automated cars on the same roads shared with human drivers is potentially dangerous as humans cannot fully predict a computer’s actions, and computer logic may fail when faced with illogical humans.
However, policy should be designed to foster technology, not hinder it; over time we would gain much safer road conditions. Policy makers have already clashed several times with autonomous vehicle companies due to their lack of knowledge – last year, California banned the testing and deployment of Google’s driverless cars as they did not trust them to be fit for the road without enabling human intervention in case the system failed. This turned out to be a policy based on wrong grounds, as scientific research actually demonstrated that “the assumption that humans could be a reliable back up for the system was a total fallacy”. Going forward, policies should be made based on thorough understanding of the technology.
Ethics is also a burning issue, illustrated by the theoretical question about how the car should be programmed to act in the event of an unavoidable accident. While it could make sense to value the passenger’s life more than a pedestrian’s life, the ethical considerations are tricky. Automobile companies will have to find the correct answer to this kind of questions (and many others) for social acceptance.
Cars are seeing huge changes, and the industries that are tied to cars will undergo significant shifts as well. Both incumbent car companies and new players, as well as companies supplying these with parts and solutions have big opportunities, as long as they take on innovation and look with open eyes what technology can provide. With Tesla and other innovators like Faraday Future entering the field, things are moving faster. Many automotive companies are establishing permanent presence in Silicon Valley to stay on top of innovations. Even traditional players like BMW are surprising the consumer with highly innovative products through their ‘i’ series electric cars. The “genie is out of the bottle” and opportunities loom around for those fast enough to capture them. Expect to see big changes in industry structure and value chain too – the automotive industry of the next ten years may be led by some new players and new solutions.
Links for previous article series:
(1) Public Cloud’s Growth and Outlook – Addressing New Opportunities
(2) Leveraging Social Networking Services with the Power of Big Data
(3) Mobile Payments – The Battle for Customer Insight
(4) Virtual Reality and Augmented Reality: Upcoming Consumer and Business Powertool
(5) Indoor Tracking and Maps: The Hype Died, but the Show is Not Over
(6) Digital Signage – Entering the Personalized and Interactive Era
(7) Artificial Intelligence – Tailored Smart Solutions Today, True Intelligence Tomorrow
(8) Digital Health – Disrupting Healthcare Technology and Democratizing Health Care Services
Written by Per Stenius, LG CNS Blog Partner, and Seoweon Yoo
Further reading and references:
This blog is based on a broad range of articles, books and reports. Some of the more interesting ones are listed below.
 Travis Kalanick, Uber’s plan to get more people into fewer cars (March 2016), Ted.com https://www.ted.com/talks/travis_kalanick_uber_s_plan_to_get_more_people_into_fewer_cars/transcript?language=en [back to the article]
 https://www.waze.com/ [back to the article]
 https://play.google.com/store/apps/details?id=com.locnall.KimGiSa&hl=ko [back to the article]
 https://www.uber.com/ [back to the article]
 http://lyft.com/ [back to the article]
 http://www.wired.com/2016/05/calling-uber-cooler-owning-car-automakers-want/ [back to the article]
 http://www.zipcar.com/ [back to the article]
 http://car2go.com/ [back to the article]
 http://www.apple.com/ios/carplay/ [back to the article]
 https://www.android.com/auto/ [back to the article]
 http://www.qnx.com/ [back to the article]
 http://www.recode.net/2016/1/14/11588804/the-hottest-computing-device-cars [back to the article]
 http://www.forbes.com/sites/lianeyvkoff/2015/11/02/king-of-infotainment-qnx-isnt-afraid-of-google-but-it-should-be/ [back to the article]
 https://www.technologyreview.com/s/534981/car-to-car-communication/ [back to the article]
 http://www.wired.com/brandlab/2016/03/a-brief-history-of-autonomous-vehicle-technology/ [back to the article]
 http://www.novatel.com/industries/autonomous-vehicles/#technology [back to the article]
 https://www.teslamotors.com/blog/your-autopilot-has-arrived [back to the article]
 Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future, Ashlee Vance, HarperCollins Publishers (2015) [back to the article]
 http://www.mobileye.com/en-us [back to the article]
 http://www.forbes.com/sites/joannmuller/2016/06/15/mobileye-is-the-auto-industrys-secret-weapon-against-google-in-the-race-for-self-driving-cars/ [back to the article]
 http://www.reinventingparking.org/2013/02/cars-are-parked-95-of-time-lets-check.html [back to the article]
 http://www.reinventingparking.org/2013/02/cars-are-parked-95-of-time-lets-check.html [back to the article]
 https://www.teslamotors.com/blog/your-autopilot-has-arrived [back to the article]
 http://www.rightware.com/ [back to the article]
 http://wikispeed.org/car/ [back to the article]
 http://www.trucking.org/News_and_Information_Reports_Driver_Shortage.aspx [back to the article]
 http://safety.trw.com/wp-content/uploads/2016/01/AutomatedDriving_table_large.jpg [back to the article]
 https://www.theguardian.com/business/2016/may/03/driverless-cars-dent-motor-insurers-volvo [back to the article]
 http://www.theverge.com/2016/6/30/12072408/tesla-autopilot-car-crash-death-autonomous-model-s [back to the article]
 https://www.wired.com/2015/07/hackers-remotely-kill-jeep-highway/ [back to the article]
 http://www.computerworld.com/article/3081467/car-tech/securing-your-car-from-cyberattacks-is-becoming-a-big-business.html [back to the article]
 https://www.technologyreview.com/s/542651/drivers-push-teslas-autopilot-beyond-its-abilities/ [back to the article]
 https://www.teslamotors.com/blog/tragic-loss [back to the article]
 http://www.forbes.com/sites/chunkamui/2015/12/18/california-slams-the-brakes-on-googles-driverless-car/#528f5645e8c1 [back to the article]
 http://www.theverge.com/2015/3/17/8235277/sxsw-astro-teller-google-x [back to the article]
 Autonomous Vehicles Need Experimental Ethics: Are We Ready for Utilitarian Cars? http://arxiv.org/pdf/1510.03346v1.pdf [back to the article]
 http://www.ff.com/ [back to the article]
 http://www.wsj.com/articles/ford-mercedes-set-up-shop-in-silicon-valley-1427475558 [back to the article]
 http://www.bmw.com/com/en/newvehicles/i/overview.html [back to the article]