Automated Cars Fall 2015

From New Media Business Blog

Jump to: navigation, search

Chris Urmson’s TED Talk: How A Driverless Car Sees The Road?
Automated cars have become a reality since 2009 when Google launched the first Google Car. It came a market reality in October 2015 when Tesla updated all its cars with their Autopilot feature.

It is raising huge interests for companies, for citizens and the whole world in general. Chris Urmson, the Director of Self-Driving Cars at GoogleX, explains very well in the first few minutes of his TED talk why the technology is interesting today.

Very quickly, one can see the amazing possibilities and opportunities in one’s everyday life if automated cars were widespread. It would save costs & time for individuals and companies. It could also reduce traffic, pollution, health problems, etc. The disruption in the car industry would revolutionize the entire economy and the entire way of living since it’s been a leading sector since the 2nd industrial revolution in the XIXth century. Some new challenges could arise are: safety, regulation, insurance, lobbying, energy, geopolitics, employment, behaviors, etc.



There are a variety of name being used throughout industry for the developments in automated car technology including: automated, autonomous, driverless, self-driving or even unscrewed or robotic. The definition is already discussed on the Autonomous Cars Wikipedia page [1].

From governments' point of view, some started to make a distinction between levels of automation, like in the US. The classification system proposed by the National Highway Traffic Safety Administration (NHTSA) is the following: Level 0: The driver completely controls the vehicle at all times. Level 1: Individual vehicle controls are automated, such as electronic stability control or automatic braking. Level 2: At least two controls can be automated in unison, such as adaptive cruise control in combination with lane keeping. Level 3: The driver can fully cede control of all safety-critical functions in certain conditions. The car senses when conditions require the driver to retake control and provides a "sufficiently comfortable transition time" for the driver to do so. Level 4: The vehicle performs all safety-critical functions for the entire trip, with the driver not expected to control the vehicle at any time. As this vehicle would control all functions from start to stop, including all parking functions, it could include unoccupied cars.

From users' point of view, what needs to be remembered are the last 2 levels that one could call semi-automated cars (level 3) and fully automated cars (level 4), because they are the current reality of the technology development. First levels (0, 1 and 2) already exists and are on the market, while level 3, semi-automated cars technology, has just been implemented by Tesla (and Freightliner in the trucking industry). On the other side, fully automated cars are the technology publicly developed by Google, which hasn't yet publicly announced how the company would hit the market with it. Tesla and Google are the 2 current public leaders concerning the technology: Google is the technology pioneer and Tesla is the market pioneer.

Tesla's Autopilot (semi-automated cars / level 3): the car can drive itself if the driver choses so. There is a steering wheel but the users should keep theirs hands on it. The technology was announced in 2014 for 2015 and it did come out in 2015, as opposed to BMW, Mercedes or Volvo who have made announcements, but are yet to get a product to market.

Google's Driverless Car (fully automated cars / level 4): the car completely drives itself. It has already driven several millions miles, collecting data. There is no steering wheel, just an emergency brake button. There is not yet a business model publicly announced by Google for their fully automated car.

The 2 technologies are closely linked because Elan Musk (CEO of Tesla) predicts that fully automated cars will be on the market around 2020 [2]. And this is very important to keep in mind because the implications of these 2 technologies are very different for all the topics discussed below.


The Linrrican Wonder, 1925
The concept of autonomous cars has been around since the 1920s and has since gone through major technological advances. The first autonomous car was a radio-controlled driverless car called “Linrrican Wonder” from Houdina Radio Control which traveled along the streets of New York in 1925. Francis Houdina had used a 1926 Chandler model and mounted antenna on top of the car so that it received radio signals from a second car that followed it. This way, the second car operated the driverless car via radio transmission through the antenna. [3] The following year, Achen Motors, a distributor of cars showed a demonstration of autonomous cars in Milwaukee under the name “Phantom Auto.” [4]

After these inventions, 1950s and 1960s saw a lot of trials in regards to autonomous cars as well as infrastructure research. One of the prominent ones were the RCA Labs in the 1950s who collaborated with GM Motors and State of Nebraska to build an autonomous car that operated through wires buried inside an experimental detector circuit underneath the pavement of a 400-foot strip of public highway in Nebraska. The wires guided the motion and the direction of the autonomous vehicle. [5]. Then in 1960s, Mercedes Benz along with the Bundeswehr University Munich in Germany created a robotic van that was vision guided and was driven at a speed of 39 miles per hour without traffic. [6]. In 1960s, Ohio State University’s Lab worked on a project to build driverless car that could be activated by electronic devices buried underneath the roadway. Similarly, UK’s Transport and Road Research Laboratory worked on a project to test driverless cars, 1960 Citroen DS, that were activated by magnetic cables underneath the road. However, due to financial issues, funding for the project were withdrawn by the 1970s. [7]

In the 1980s, a major breakthrough took place with the Autonomous Land Vehicle (ALV) project in the US. The project was collaboratively done by University of Maryland, Carnegie Mellon University, SRI International and the Environmental Research Institute of Michigan. The project was the first one to use lidar, computer vision and autonomous robotic control to direct a robotic vehicle with speeds up to 19 miles per hours. By 1989, the use of neural networks to control autonomous vehicles as well as the use of road maps and sensor based navigation to steer the vehicles had been achieved through the ALV project. In fact, these are the same control and navigation system that form the basis of automated cars today. [8]

From 2009 to today, there have been 25 companies around the world that have publicly announced that they are working on autonomous vehicles. Out of the 25, the most famous one is Google because it triggered the revolution in 2009 with its first driverless car that has been evolving ever since. The second most famous is Tesla because in October 2015 it actually released the AutoPilot, which was first announced in 2014. We can divide these companies in 3 categories : those doing research on fully automated cars with an average release date around 2030, those doing research on semi-automatic cars with an average release date around 2020 and those developing technologies and features (such as GPS tracking) [9] It is evident that the automated vehicle technology has evolved and advanced throughout the years, however, there are issues surrounding user behaviour, safety, infrastructure, regulations and business implications that has prevented the autonomous vehicles from being market launched.

User behavior

The first topic that needs to be tackled with automated cars is the user behaviour because the drivers will be one of the main pillars of the technology’s development. From a design approach, understanding the reasons behind driving and the reality of the drivers may help in finding the right way to develop the technology, i.e the debate between semi-automated and fully automated cars. Even if a lot of studies have been done on the public opinion about automated cars [10], one still need to go deeper in the drivers’ behaviour to understand the conditions with which the technology could be developed.

Driving habits

The large majority of the driving is commuting to and from work. 85% of the Americans drive to work (the other solutions being taking public transportation, biking or walking), and most of the time, they are alone in their car. So one could say that traffic jam is due to the lack of optimization on how the people use their car. Carpooling is still minority behavior despite the development of the sharing economy and carpooling platform like Blablacar in Europe [11]. 82% of the European population drives less than 100km a day. Usually, these small trips take place in the city centers or the suburbs where there is a lot of traffic.

However, some initiatives showed that this habit of car commuting, i.e. associating of driving and going to work, can be easily changed in a city with good public transportation. WWF UK managed to establish new behavioral patterns after moving its headquarters in a new city [12]. Taking advantage of this major event, the non-profit completely changed its incentive system on transport by reimbursing the difference between driving and taking the train. In one week, after the relocation, car commuting decreased from 55% to 29% of the employees and train commuting did the opposite, rising from 18% to 56% of the employees. So what it shows regarding the development of the automated car technology is that people wouldn’t mind using alternative transportation to go to work, that is to say for a majority of their driving time.

Customer Value proposition

Looking at this potential disruption with a business approach can be interesting. Instead of studying what would deter people from using automated cars, one can also take the angle of studying what would convince the users. In general, Rogers showed [13] that the speed with which an innovation can spread on a market depends on its advantages over existing solutions, its compatibility with existing values, its user-friendliness, its triability and its observability.

Going a bit further on the automated car innovation in the car industry, the good rationale according to KPMG[14] is hit the market with the following arguments: shorter commute time, reduced traffic-related variability and freedom to turn self-driving mode on or off. This is what they call the right value proposition for the customer. Again, the first argument has something to do with work. So the convincing selling argument is on the everyday driving inconveniences.

Nevertheless, there are two examples in the history of the car industry that shows that the very way companies make business in the car industry is also unpredictable. The first example is Tesla, which was the first company to hit the market with an semi automated car. Tesla first sold premium electric cars with much higher autonomy range than the mass-market average (400km compared with 130km). Then the automation of the car is an software update. The approach is completely different from what one could imagine about selling cars. On the one hand, the car is now a computer and the automation is a piece of software, which makes car makers service providers. On the other hand, the automation was free, which means that there is no premium to have access to automation for now. The other example is the failure of GM to sell electric cars in the 90s. From a business point of view, the main reason why it wouldn’t work was that people wouldn’t pay more for a car than can do less [15] although the car was better for the environment and was matching the every day use/need of most of the people. So the main lesson on the car industry in general here is that it is very difficult to know what represents “more” for the consumers, what is really an advantage for them and foremost, to what extent would they be ready to pay more, if anything, for automated cars.

Psychological Barriers

From a psychological point of view, according to Lucas, Blumenberg and Weinberger (2011) [16], “Our car use behaviors are largely a function of factors way beyond the standards of time and cost”. There are indeed irrational or social obstacles if automation means that people won’t be able to drive anymore. One of them is the pleasure of driving (I can accelerate and I like being in control). It is not necessarily contradictory with automated driving because, according to the KPMG study mentioned before, people keen on driving as a pleasure are the one most easily convinced by automated cars. Another obstacle is the pleasure of mobility (I can move when I want). As long as people can have access to a car, owning it is not necessary. This is a possible vision of Uber [17]. Social status could be an obstacle because owning a car is a social standard and a synonym of social integration. Nonetheless, there are generational differences. The current trend is that mobility becomes the main priority over ownership. The development of the sharing economy is the best evidence for that. Car sharing is booming (Autolib in Paris, France: Car2Go or ZipCar in Vancouver, Canada).

Trust in Artificial Intelligence

The trust in machines is not just a social problem, it is also ethical. On the one side, the trust in electronics is quite irrational because, even if electronics are not perfect, using it for cars would still make the cars safer than a human driver. Some studies even showed that making machines more human can make people more trustful with machines[18]. On the other side, there is an ethical problem [19]: how can one program a machine to make a decision that people don’t make themselves, when in a car accident for example? Does one even have to program the car for that kind of situation? Is this question even relevant, meaning how much do the situations described in the article represent? In the end, ethics questions links perfectly user behaviour obstacles and safety obstacles.

Safety concerns

One of the main objectives of driverless cars is to establish safety to its passengers. But paradoxically, Goodyear surveyed young drivers, aged 18 to 30, about the rise of autonomous vehicles and only 11 per cent said they would have full confidence in getting in driverless cars [20]. These figures show that safety remains a critical challenge among car industries, and this is true in several domains.

Safety on the road

Safety on the road remains the main concern regarding driverless cars, which hides a great paradox. Theoretically driverless cars would significantly decrease the risk of accidents. However, the public opinion remains reluctant to this technology and still percieves it as dangerous.

The paradox of safety and driverless cars

Besides the psychological reason, so far, statistics and figures justify this position from the public's opinion. Indeed, even if theoretically in a world full of driverless cars accidents cannot happen, our reality is different and driverless cars have to share the road with regular drivers. In this context, driverless cars are involved in car accidents, which confirm the potential risk of driverless cars on the road. So far, 16 Google driverless cars have been involved in accidents for instance [21]. Statistically, self-driving cars are five times as likely to crash than conventional ones, and their passengers were four times as likely to get injured (with 3.29 injuries per million miles, compared to only .77 injuries in regular cars) [22].

This could be explained by the fundamental differences that exist between driverless cars driving style and human driving style. Thanks to technologies like radar, LIDAR, GPS, Odometry or even computer vision, driverless cars are able to "see" their surroundings, analyze the environment and identify appropriate navigation paths. Driverless cars' vision exceeds human perspectives. As a result, driverless cars appear too cautious and prudent in the eyes of regular drivers, who can't always understand why a driverless car decided to break. All the accidents reported today concerning driverless cars imply a rear-end collision because a human driver did not react in time after one driverless car's braking.

That's why, with the safety issue comes the following challenge: how can driverless cars adapt to human drivers? How can a driverless car act more like a human being on the road for safety reasons? Car industries through trial and error attempt to alter artificial intelligence into common human responses. Another solution would be to integrate different profiles inside driverless cars in order to better fit their passengers needs and wants [23].

Health safety concerns

The development of driverless cars is done in parallel, with the development of electric cars. We can assume that in the future, driverless cars would be mostly electric. However, electric cars prevails its own set of issues, the electro-magnetic field exposure which could potentially be dangerous according to several studies. According to these researches, EMF exposure could increase the risk of cancer, leukemia, anxiety and cause headaches.

But these concerns have been disclaimed by contradictory studies. The truth is that nothing has been literally proven around the EMF exposure risks [24]. It is an ongoing debate, which also concerns high-voltage lines, mobile phones, and now driverless cars, especially because cars exposure are usually long.

Even if EMF exposure is not vital risk for health, it is a concern, which needs to be addressed. Indeed, all potential safety concerns should be resolved in order to fully implement driverless cars on a large-scale. Car industries can use Faraday cage around cars' batteries. This would ultimetaly eliminate all potential risk, and conclude this debate.

Hacking and driverless cars

Driverless cars are high-technological products, and involve numerous technologies that can be attacked by hackers.

Hackers Remotely Kill a Jeep on the Highway—With Me in It

Even non-professional hackers can trick a driverless car, thanks to a 60$ handheld laser [1]. All comes from the LIDAR (Light Detection and Ranging) system than enables most of the driverless cars to visualize and analyze its environment to react properly. A simple handheld laser can trick the car into thinking there are objects where there actually is not. Everyone is so able to force a driverless car to slow down, stop or swerve.

The risk is even worse when it comes from professional hackers, who with the wifi access are much more skilled and capable to manipulate driverless cars, as was the Jeep some hackers exploited to kill remotely when it drove down a highway. Fiat Chrysler Automobiles had to recall nearly 8000 SUVs after this incident.

Here, the risk is important, especially because the potential consequences of a massive attack can be serious: hackers can take the control of a specific car to kidnap or kill ; hackers can also paralyze the traffic and an entire city.

Cars industries have to take measures in order to prevent this important safety issue. Especially because nothing is really hack-proof. Hackers would ever be able to attack driverless cars. The only way to deal with this risk is to elaborate a cyber-security strategy that fits hackers' wants and ambitions [2]. So far, hackers do not seem interested in hacking driverless cars, probably because this technology is still anecdotal. However, with the future improvements concerning the implementation of driverless cars, cars industries in association with government and cyber-security companies should remain cautious.


Country – major highways/freeways

The first distinction has to be made for highway and freeway. Among other things, a freeway has a divider between on-coming lanes of traffic, while a highway does not. Meaning there is an actual barrier (concrete, grass, space) between oncoming lanes of a freeway, while there is only road lines between oncoming lanes of a highway.

In general the major freeways of the Western World are already well set up for the auto-pilot features in operations today. Both Tesla’s and Freightliners auto-pilot are designed around being used on the freeway, with all traffic moving in the same direction. Therefore, there is little to no infrastructure changes needed to freeways, to allow fully autonomous vehicles from operating here.

For highway, we would need to build, or somehow divide oncoming lanes, in order to allow the existing technology to operate in these environments. This is unlikely to happen on mass scale, it is far more likely the software involved in the vehicles will adapt to allow for highway driving of fully automated vehicles.


NASA SkyTran

If we move towards full automation, it will mean our existing rules of the road will continuously be challenged. In our utopian world, traffic movement will be seamless, however the realities we will face while working towards this utopia will be difficult and numerous. It is unlikely that infrastructure will adapt as quickly as the technology, as the cost of doing so is simply too great. Overtime infrastructure must adapt, but it’s impossible to anticipate the effects of automation on this adaptation.

Google’s fully automated car is currently best suited for suburban and city areas. As its speed limit is capped at 25mph it isn’t well suited for highway driving, however it is perfected suited for short commutes through city or suburban areas.

On November 23, 2015 NASA released and article stating their patented “SkyTran” technology is closer than ever. The technology uses and overhead magnetic levitation railway to power 2 passenger pods. Magnetic levitation makes it very efficient, and the cost of building the structures is fractions of the price of current subways, or overhead tram lines (Second Nexus).

While this sounds great, reality is it would take a huge cost to build these overhead tram lines, and because they only hold 2 people, it likely means we will need more of them in our major cities and surrounding suburbs. The concept is great, but the realities of how it plays out remain to be seen.

City centers

It is estimated that 50% of the land in most major cities of the world is taken up by roads and parking lots. As our city centers continue to get denser and traffic congestion becomes more intense cities look to solve this problem.[3]. Two major cities in Denmark, Copenhagen and Aarhus, are already well known around the world for their promotion of bicycles. Aarhus is currently experimenting with using RFID chips attached to people bicycles. When a cyclist approaches a stop light, a sensor reads the RFID chip and turns the light green, stopping car traffic in the process. Other major cities use regulations to restrict driving in their dense city centers. Cities charge fees for licenses, cars, or special permits to operate in certain areas, in order to limit the amount of vehicles that operate in specific areas. London is one example, where you need to pay a fee to operate in Level 1 (central London). It is no doubt that regulation will play a major role in the adoption of automated cars operating in major cities of the world. If mass automated transportation become viable, it could be possible for a city to permit the use of automated cars only, maximizing efficiency, and therefor freeing currently used road and parking space. Automation could also allow the possibility to “send your car home”. Which would mean parking in major cities could become obsolete. If we take our utopian assumptions into consideration, it could be possible for a company like Uber to own and operate it the entire automated fleet of New York City taxis. This has massive implications for efficiencies and infrastructure.


As driverless cars start rolling onto public roads for testing, there have been some developments around the world in terms of regulations. It is evident from the way governments have reacted so far, that new laws will have to be mandated before driverless cars can be market launched. This means new legislation around not just driving but also around liability, which ultimately affects how fast the driverless cars will be market launched. Before stepping onto the future, it is important to know which parts of the world have allowed road testing of automated vehicles on public roads, the discussions on insurance and liability so far and finally, where might the launch be happening first.

Road testing

US states that have allowed autonomous vehicle road testing
In the US, Nevada was the first state that allowed testing of automated cars on public roads in 2011 with the requirement of a driver’s license endorsement for operators of such vehicles. The state’s Department of Motor Vehicles (DMV) were the ones to adopt rules for license endorsement and for operation, including insurance, safety standards and testing. The DMV requires that there be persons in a legally operating autonomous vehicle behind the wheel and in the passenger seat. The DMV has permitted use of wireless handheld devices for those persons since they are not the ones driving the cars. The state worked with Google to bring about this legislation. [4]

Other states in US that have allowed testing of automated cars are California and Florida in 2012, Michigan and Washington D.C in 2013 and North Dakota and Tennessee in 2015. All of these states have their Department of Transportation or Department of Motor Vehicles that set rules and regulations for the testing. They require that persons be present in an autonomous vehicle and that the operators get a license before testing the vehicles on road.[4]

In France, PSA Peugeot Citroën, the second largest car manufacturer in Europe, received authorization to carry out open road tests for four of their autonomous models. They completed a 360 mile trip from Paris to Bordeaux in October 2015 after receiving the authorization back in July. A driver was required to be present in the driver seat during the trip with their hands near the steering wheel in case of an emergency.[5] Similarly, state authorities in Germany have allowed Daimler to test drive their autonomous cars in Baden-Württemberg in September 2015. The state requires that there be a driver in the car for the testing to take place. [6]

The Lutz Pod
On the other hand, the government of UK has published the following revised guidelines in 2015, becoming the nation with the world’s first code of practice for autonomous vehicles:
  • A minimum of 30 seconds of data must always be available so if the driverless car is involved in an accident, the cause of the accident can be determined;
  • The pavement 'pods' like the Lutz Pod, which travels on pavements and part of the first driverless networks, must have someone who can remotely control them (but doesn’t need to be inside them) to bring them into a safe state in the event of a problem ;
  • The driverless vehicles called M1 must have a manual driver in the vehicle who can take over if the autonomous system fails;
  • If there is an accident, in the case of the M1 vehicle the liability lies with the driver, in the case of the driverless pods it lies with the remote controller of the vehicle.

As of now, four towns and cities: Greenwich, Bristol, Coventry and Milton Keynes have authorized testing of autonomous vehicles on public roads. [7]

In Canada, Ontario became the first province to authorize test driving of autonomous vehicles on public roads. The testing of automated vehicles will begin as early as January 1, 2016 with the autonomous vehicles having to be licensed and apply to be registered with the Ministry of Transportation. A copy of the permit will need to be kept in the vehicle at all times and produced if stopped by a police officer. Those participating in the pilot program will have to have a minimum liability of $5 million.[8] In order for the vehicles to be allowed on roads, there must be a trained driver who knows how to operate an autonomous vehicle present in the driver’s seat at all times so that the driver can take over if anything happens to the system. The autonomous vehicles will be subject to the speed limits and other regulations that exist under the province’s Highway Traffic Act.[9]

Insurance and Liability

Once these vehicles are testing on the road, accidents are bound to happen so the question of who is liable and who needs to be covered is becoming a worldwide issue. Volvo has accepted full liability in the case of an accident if it is caused by a flaw in the car’s design. However, if the customer misused the technology, then the user would be liable and if a third party is the cause of the accident, then the third party would be liable in the case of an accident.[10] Similarly, Google and Mercedes Benz are said to be taking the same stance as Volvo when it comes to liability issues. [11] On the other hand, Tesla Motors Inc. has said that the driver is liable in the case of an accident because their semi-autonomous vehicles require the user to switch the self-driving mode on. [12] Other automakers have not taken any stance so far, and it is perhaps that they are waiting for the regulators to make the law unlike Volvo and Tesla.

Theoretically speaking automation would lead to reduction in accidents, which in turn, would lead to reduction in insurance premiums. However, with the recent announcements of automakers taking responsibility for accidents, the insurance premiums might see a shift from drivers to manufacturers rather than reductions. However, if the liability does shift from drivers to manufactures, then UK’s laws would have to be changed as currently it doesn’t recognize this approach to liability. [10] Similarly, the laws around the world would have to be rewritten if autonomous cars are introduced. For example, in Canada’s B.C province, the liability of a vehicle doesn’t just lie with the driver but also with the registered owner of the car.[13] If automated vehicles are introduced, then this adds one more player, the manufacturers, to the the mix complicating matters even further. If the automated cars bring in a shared ownership economy, then the liability issue becomes even more muddled.

Market Launch

Most of the governments around the world have poured in investment into research and development of autonomous vehicles. For example, in Canada, Ontario has offered $2.5 million to support project that encourage of innovations for automated and connected vehicles through its Connected Vehicle/Autonomous Vehicle. Likewise, UK government has pledged £100m over the next five years into research and development of automated cars and supporting systems they require.[14] Even though governments have made investments into development of autonomous vehicles, there are still unanswered questions around ethical issues such as privacy.For example, what happens when an autonomous vehicle becomes the subject of a crime? Can police officers access entire data on the vehicle for the purpose of the investigation? Some are wondering if the investigation of an automated vehicle would require a search warrant or if it would be treated like a cell phone case where warrantless search is permittable. [15] Another gray area would be regarding age restrictions for driving an autonomous vehicle. Should the government regulate a law to allow kids to drive an autonomous vehicle? Will there be any age restrictions or limits on who can and cannot drive an autonomous vehicle?

There is definitely a long way to go before regulations are formed and implemented that would allow autonomous vehicles to be market launched. Delay in the launch of such vehicles are what the automakers around the world want to avoid. In 2014, the UN Convention on Road Traffic was amended to allow autonomous driving if a person in a car is still able to shut the engine off. All the EU countries have signed onto this convention, but the United States, where large tech companies are testing their own autonomous vehicles, is not a signatory. Due to this, many are wondering if Europe is going to be the hub where the autonomous vehicles will be launched first. Regardless, there is no doubt that the technology is moving faster than the regulators can respond. As of now, automakers can keep testing their autonomous vehicles and onlookers can keep their eyes out for any new developments from the law as well as from the automakers.

Business Implications

Consumer Adoption

There are a diversion of opinions about which direction we would head to with autonomous vehicles and how long will the process take. The most aggressive bulls on self-driving cars see the first fully automated cars to be sold in the market in 4 to 5 years with a steady market penetration, while some other companies hold a more conservative thought on how long the process will take. It is important to note that most of these bulls are not from the traditional auto industry, and the general expectation of having autonomous cars on sale is at least 10 years away among auto Original Equipment Manufacturers(OEMs) and suppliers . Morgan Stanley has given their version of timeline for adoption, within which there are four phases: passive autonomous driving, limited driver substitution, complete autonomous capability and the utopian society. In this timeline, we will be entering the first phase where autonomous capacity is simply a second line of defense in less than 3 years; and in two decades, we are hoping to see a world in which all cars on the road are capable of driving themselves without any human intervention. [16]

Timeline for adoption by Morgan Stanley

However, some industry researches made the argument that the switch over to autonomous vehicles could take much less than several decades. The reasoning behind their claim is that there’s a compelling business nature of self-driving cars: they can make money for the owners by moving humans and goods. In fact, they would be the first autonomous robot that the masses can own and use to work, and such profit generating potential could definitely lead to a pent-up demand for automated cars. Another reason is that the transportation-as-a-service principle is likely to leverage the market value of autonomous cars, meaning that a single, fleet-owned automated taxi can effectively replace numerous private vehicles.[17]

Automotive Industry

Despite the fact that autonomous cars might reshape the industry, we must keep in mind that the number of automobile manufacturers in the U.S has decreased from more than 1500 in 1900 to fewer than 50 by 1950. For decades a few large U.S automotive manufacturers dominated the competitive landscape; now they have to fight for market share with major European and Asian automakers, along with a handful of emerging startups such as Tesla.[18] Many researches indicate that the shift in the competitive landscape might have just begun.

The emergence of autonomous cars can potentially provide a new paradigm to the auto industry in two major ways. The first is the emergence of software as a key part of the value of a car; the second one is introducing a new revenue model by being able to monetize the content opportunity within the car. Compared the traditional industry where the OEM is the most important link of the supply chain and contribute most of the content, the autonomous car industry has a more complex structure consist of hardware, software and content providers. The software provider will play a big part in the industry by enabling fully autonomous capabilities and that the value of software in the car will certainly increased. Content for entertainment, productivity or functionality in the autonomous vehicle can also be seen as an value-added product in the new auto industry.

The auto OEMs will be affected the most by the move to autonomous cars. In order to succeed in the new auto industry, the OEMs will have to assure the reliability and feature-rich autonomous capabilities of their cars. Moreover, if we get to a world where autonomous cars become widely accepted and non-autonomous cars are no longer allowed on roads, OEMs who cannot guarantee the two key elements will either be forced to go out of business or reinvent their business models to be hardware suppliers for autonomous car producers. It is also important to point out that getting a head start on testing and development of the self-driving vehicles is critical for companies to succeed in the new auto industry. The reason behind this is that there is a significant amount of effort and time needed to ensure an autonomous vehicle’s reliability and safety assurance, and this is something that can only be gained with experience. Although it might be tempting to wait for other firms to clear the path for the first few years, gaining a second mover advantage or trying to be a rapid follower is likely to be very expensive or nearly impossible for late entrants in this case if they are caught late 5 to 7 development years behind their competitors. However, it might be much harder than it sounds to have early success in the new auto industry, with various firms such as General Motors, Ford, Toyota all making big moves towards self-driving vehicles, and facing threats new entrants such as Tesla and Google. Not to mention that OEMs will need to balance and prioritize their resources on a variety of technologies including fuel efficiency, safety and infotainment technology.

Impact on Other Industries


Morgan Stanley estimated that fully autonomous cars could free up about 75 billion hours of time that are currently spent by drivers each year, which equals to 6 to 7 hours per week per driver. The implication of that is there could be a potential materially increase in the total media consumption due to the leisure time that will be freed up. However, not all media will be the beneficiaries, since there has been a shift from pure audio consumption to video-based media. A large portion of current drive time is spent listening to audio, most notably radio, and it is very likely that the total radio and recorded music listening time will be at risk of transitioning to other media with self-driving cars become available. In addition, by significantly expanding the opportunity to consume video, autonomous cars could potentially made TV the largest beneficiary on a total dollar basis and home video to benefit the most on a percentage basis.

Car Rental

It is obvious that autonomous driving will have a great impact on the car rental industry, but the direction of the impact can be very positive or negative. We can think of highly polarized scenarios ranging from a world in which self-driving vehicles increase the benefits of private ownership and usage, cannibalizing the need to rent vehicles, to one in which a roving parc of public transportation vehicles is controlled by firms with fleet management and customer service expertise. On top of that, it is very likely to see lowered entry barriers for car rental firms if vehicles become connected devices and labor cost can be greatly reduced. New business models such as peer-to-peer car rental will become much easier to implement, hence impacting the size and share division of the car rental pie. Nonetheless, while we believe that autonomous cars will encourage car usage by eliminating hazards of human driving, there is also great uncertainty over its impact on private ownership.

Freight Transportation

It is believed that autonomous and semi-autonomous driving technology will be adopted far faster in the cargo markets than in passenger markets. For the reason that people are far more comfortable with autonomous vehicles operating when human life is not at risk. In fact, autonomous driving system has already been put in use in many non-passenger environments. For example, the dump trucks in Australia mines, military truck convoys in war zones and drone military aircraft. When focus on the trucking industry, autonomous vehicles can bring potential savings in a lot of areas. Fuel efficiency saving could be fairly large since fully autonomous trucks can be put on cruise control and carriers could create a train of rigs on the highway to lower air resistance. Saving of productivity comes from many forms such as more efficient route planning and less congestion. But it is difficult to give comprehensive estimate on the productivity gains from the adoption of autonomous driving technology since we must take into consideration the maintenance and cost of the trucks and many other factors. Another gain from adopting self-driving trucks comes from accident savings. By using autonomous trucks, transportation companies can potentially avoid human-related accidents and many lawsuits associated with them.

Autonomous Truck
Perhaps the most contradictory gains will be labor savings. Carriers view the driver as a cost, one that must eventually be returned to home base. As a result, most investors may believe that the concept of autonomous freight vehicles is a certain way to reduce or eliminate labor costs, which are the largest cost bucket for any freight carrier. However, it is questionable that whether the labor component will be fully eliminated. Although the number of drivers required in an autonomous driving environment will decrease, labor will be needed for programming, route planning, maintenance and fleet managers. The real saving for carriers could come from fleet productivity once trucks can run by themselves 24 hours 7 days a week.

Google and Its Approach

Google’s self-driving car program consists of around a dozen vehicles, primarily Toyota hybrid vehicles, each supervised by a driver and an engineer in the front two seats. Google has successfully sponsored legislation legalizing self-driving cars (with human caretakers) in three US states: California, Florida, and Nevada. As of August 2012, Google’s autonomous vehicles had driven more than 500,000 kilometres with no accidents. According to various media sources, each of Google’s vehicles contain around $150,000 of computer and sensor equipment, including a $70,000 LIDAR system that measures objects and distances via a combination of laser imaging and reflected light. Although the expensive hardware and software of the prototypes might make it hard for Google to introduce its self-driving cars at a affordable price for the mass, several Google management have stated that the firm is confident in bringing a reasonably priced system in the nearby future. Furthermore, Google could potentially penetrate the auto market in a number of ways. The first one is using robotic vehicles to collect precision location data, Street View data, and business data. Secondly, Google could choose to sell or license its autonomous software system to traditional vehicle OEMs, since it has its advantage in gathering consumers' search data and understanding consumer behaviour. However, this is under the assumption that the OEM is confident about autonomous cars and is willing to invest heavily in autonomous vehicle production. Alternatively, Google might decide to develop its own self-driving cars and sell them to customers.Also, Google may believe that computerized drivers are more likely to choose efficient routes and minimize fuel consumption. Google may see these as public goods, but ones that it can most effectively deliver by producing a fully integrated hardware/software vehicle. By now, it seems that the last option is the most promising one. Mashable's article on November 30 this year indicated that Google has recently hired Tesla's engineering manager Robert Rose, whose expertise is in converting human-driving cars into semi-automated cars. Although it is still not clear what Google's intention is, it is safe to assume that it's planning to produce and market its own autonomous cars.[19]

Driverless cars in fiction

Science fiction author often gets ahead of the real scientists and imagines revolutionary technologies. It has been the case for planes, submarines, humanoids and plenty of other examples. Driverless cars are no exception, and artists developed different worlds where driverless cars are the norm. It is interesting to compare these artists' vision with what is really happening nowadays.

Driverless cars in literature

World's fair representation

The science-fiction author Isaac Asimov is maybe the best and most famous example, especially after the film adaptation of his book I-Robot written in 1950, where he imagined a car with a manual-driving mode, which is extraordinarily similar to what Tesla implements in 2015. In 1964, Isaac Asimov wrote an essay in the New York Times after attending the World's Fair, where he imagined a visit to the World's Fair 50 years in the future that is to say in 2014. Among his predictions: “Much effort will be put into the designing of vehicles with ‘robot-brains’—vehicles that can be set for particular destinations and that will then proceed there without interference by the slow reflexes of a human driver.”[20] Most of his prediction proved accurate even if the cars do not rife suspended on compressed air as he thought.

Driverless cars in film and television

I, Robot car

Driverless cars are always often displayed in movies. And it maybe started during the season two of the iconic Batman TV series where Batman uses the Batmobile’s “remote control activator” to summon the car to come pick him up. Was it the first driverless car on our screen? What is certain is that scriptwriters do not lack of imagination. For instance with the Fifth Element (1997), Luc Besson created a world where in 2263 cars would be autonomous and could fly. In general, the opposition between semi-autonomous cars (the current Tesla vision) and driverless cars (the current Google vision) is again extraordinarily present among the different adaptations. For instance in Minority Report (2002) Steven Spielberg designed a fully automated cars road system in world set in 2054. On the opposite, Alex Proyas in his I, Robot (2004) created a semi-automated Audi. This example probably gets closest to nailing what the future will look like in terms of timeline technology and aesthetics of driverless automobiles. [19]

Our vision

We also tried to imagine our world full of automated cars, and we wanted to create an optimistic vision of such a world. In this world, all driverless cars would be electric. That's why, highways would have the capacity to recharge batteries. Technically, in this world driverless cars would not have to stop driving. Parking lots would not be relevant anymore. Concerning the traffic, as accidents would not be possible anymore and because of the decreased amount of cars thanks to car sharing, it would be fluent all the time. The roads with driverless cars would be separate from ways for pedestrians and bikes. Driving would remain a pleasure, but would be only possible on certain circuits or road.



Technology continues to improve at an ever increasing rate. While autonomous features continue to improve, price continues to decrease, and as more and more players come into the game, the advancement towards fully autonomous vehicles has potential to have major impact of a lot of industries. Technology has at times had major impacts on a variety of industries, the auto industry may be the most obvious with Henry Fords invention of the assembly line. In today’s modern assembly line, there are very few people actually involved, and the advancement of autonomous vehicles has the potential to turn car manufactures into suppliers to software companies, rather than and industry in and of itself.


Cars and Trucks can be built to last, but is it to business' advantage?

Volvo Trucks - Look Who's Driving feat. 4-year-old Sophie (Live Test)

The business of technology seems to be one of planned obsolescence, where either hardware advances to quickly that it rapidly becomes out of date (think Blue Ray DVD’s) or software upgrades (which inevitably become mandatory) made the existing hardware obsolete. Car companies are also likely guilty of this, there are some vehicles that are built to last including one of Volvo’s dump trucks, and the Toyota Hilux, but most are built to only last 5-10 years and then be scrapped for newer models. All of this is in an attempt to continue to grow businesses and in the world in which business operates today, mandatory.

Reality of our planet

The third factor we see at play in the world today is the fact and realization that we live in a finite world. There is only so much materials on our earth, and we are extracting them at what most believe to be an unsustainable rate. We see the world of technology, business and reality of our planet on a collision course for each other where something must give. In some views, the invention of the internet and the ideal that the internet has a conscience will help solve this problem. Recently an executive from Pharmceutical company Turing raise the rates of their HIV drug 5000%, from $13.50 t0 $750/pill. The internet quickly spread this word through social media, and within a month a Turing competitor (Imprimis Pharmaceuiticals) release their HIV drug for a fraction of the cost ($0.99/pill) (Associated Press). In our opinion technology certainly has the potential to make our world a better place, and that on a whole technology has improved our lives, but there are those also in the world who have not benefited, and possibly who are suffered because of it. Bill Gates, Mark Zucerburg, Larry Page and Sergei Brin has all pledged to use their wealth to try and help those less fortunate. It remains to be seen if on a whole we are better or worse off for all of these “advancements”.


Personal tools