How Google Car will reshape automotive industry

What is Google Car?

Google has been working on this self-driving car project since 2005; but even then the idea of self-driving car was not new. The idea already existed as far back as the nineteen-fifties but wasn’t possible (picture 1[1]). But serious efforts at developing a driverless car had to wait until 2005, when computer and sensor technology became advanced enough to make the dream possible.

Thus in 2005 for the second edition of the DARPA Grand Challenge consisting in an autonomous vehicle race in Mojaves desert, many universities and robotics labs participated to this challenge to test their capability to design such vehicles. If in 2004 for the first edition, no team succeeded in finishing the race, for the second edition, the Stanford university team managed to complete the race (picture 2[2]). The first step of autonomous vehicles was passed. In 2007, they completed another step by moving its playground from desert to urban area with traffic. The idea of an autonomous/self-driving car for daily use was really born.

Following this success Google launched a program, Google X, to develop their own autonomous car thanks to Google street view technology and a new software called “Google Chauffeur”. This program is led by Sebastian Thrun, former director of Stanford Artificial Intelligence Laboratory and co-inventor of Google street view. A big part of Google X’s Google car team was actually made up of members of the winning DARPA challenge Stanford team.

In 2012 after Google’s intense lobbying, Nevada authorized autonomous cars operations, launching a real scale test for Google car. The same year, Florida and California followed Nevada’s legislation on robotic cars. Google’s car first version was based on a modified Toyota Prius equipped with a lot of laser, sensors and computers (picture 3[3]). This version is still drivable as a traditional car, allowing drivers to control the car in case of emergency or software error. In 2014, Google presented the new Google car (picture 4[4]), completely self-driving. This model has no steering wheels or pedals, to make them more efficient and more fit for dealing with traffic. Google plans to make those cars available on the market in 2020.

 

(picture 1&2)

(picture 3&4)

The Google car is a self-driving car which uses sensors and software to pilot the vehicles. Right now, it is built without a steering wheel; software and sensors are then totally driving the car. It still has a start button and a red emergency stop button. Two Google engineers explained at a conference how the car is able to detect obstacles and “sees” the road. (picture below)[5]

googlecarview

Chris Urmson, one of the main tech engineers, presented a laser range founder, placed at the top of the car, as the “heart of our system”[6]. This laser produces 3D map of the road and its many possible obstacles. A camera for traffic lights, a GPS, wheel encoders (to keep the track of the position and to control it) and four radars (one each side, and on the front and the back of the car) are completing the laser.

google-car-prototype

Google has been test driving the car for a while now, both on closed and public roads. Their car has driven more than 300.000 kilometers. In addition to driving on concrete and real roads, the Google car has been traveling about 5 million virtual kilometers everyday on servers where the car is subjected to all possible situations imagined by engineers. These simulators allow the car to enhance its understanding of all situations and to adapt the way it drives. In this way, thousands of scenarios are tested in just a few hours—anything from a tree falling across the road to a cyclist cutting across an intersection, or a child crossing the street in a power wheels car.

Google Car’s features are also very specific. It is said to have a “cute” look and a friendly face. The front, which is designed with what looks like eyes and a nose, is actually made to soften any potential contact with pedestrians as much as possible.

Is Google Car disruptive or not?

A disruptive innovation can be described as new market potential driven by a new product. It’s the idea of a product that could change the way we behave. Disruptive innovations have two characteristics: They are more expensive than old products and a majority of people are reluctant to use them at first.

The Google Car will completely change society’s habits and daily activities. It is safer, according to the company’s engineers. A self-driving car should limit crashes, damages and deaths thanks to sensors and software which react faster than humans. With more than 33000 deaths in 2013 according to the National Highway Traffic Safety Administration, motor vehicles accidents remain a leading cause of death factor in the US[7]. The Google car is a disruptive innovation because it can avoid accidents. But this remains hypothetical–we can’t insure a 100% reliable system due to human factor such as pedestrians and traditional cars.[8]

The second part of a disruptive innovation is the creation of new market opportunities. The Google car will allow the creation of new industries: possible accessories, maybe customization, more electric stations to recharge the car, and so on. This advanced technology can also be used in several other fields. In fact, self-driving cars will redesign our life style by repositioning the transportation from a waste of time to another possibility to share and live new experiences.

Autonomous vehicles in general will operate a big change in the freight / supply chain sector by allowing companies to create pools of autonomous vehicles that can run all day long without interruption. That new possibility will redesign the way how companies supply their goods by road, but also how they manage their drivers, and by the way create new kind of jobs. (We can easily imagine “road controllers” that are in charge of a pool of vehicles and have to check their “vital signals” to ensure the continuous supply of the goods.)

Thus according to Christensen disruptive innovation definition[9], the Google car is a disruptive innovation as well by its capability to create new market possibilities, by the reticence that it create in the population and by its price and features[10] with respect to its competitors.

On the other hand, people could argue that because the car is not yet on the market, it is still an invention and not an innovation. We were in the same situation at the beginning of the automotive industry, cars existed before Ford introduced the mass produced T-model. Cars were luxury and useless goods, the real disruptive innovation was when Ford allowed people to buy for a reasonable price a mass produced car that could replace horses in everyday life. Only then the car became a disruptive innovation. Making the parallel with the Google car this product will be a disruptive innovation when it will be available on a large scale. Until then it remains a toy for scientists and engineers.

Which industries are involved

The technology could be used by anyone who wants to move from a point A to a point B by car. More specifically, it could be very useful for people who do not have a driving license, disabled people, drunk people, elderly people, and so on. We can even think about children using self-driving cars one day. School buses for school could also be big “Google cars”.

This technology involves different industries such as the automotive industry, taxis or shuttle, rental industry, carpooling. If we go further in the technology development, adapting it to bigger vehicles, the freight industry can be involved too.

Automotive industry is going to be involved because they will have to change their production model, they will have to adapt their industry to this new mean of transport. For example, simulations and tests proved that autonomous trucks are more efficient than current drivers for delivery, it is less time consuming, oil consuming, and safer[11]. The technology used in the Google Car could be applied to a lot of vehicles.

How much improvement can we expect from Google Car?

So how much improvement can we expect to see in the autonomous car, and how far into the future can we expect to see improvement?

Human nature itself can be natural limits because of their ability to accept and to interact with this technology. Protocols could impede all the future autonomous cars on roads. Furthermore, Google will not be the only self-driving car manufacturer and retailer. Google might put its protocols on open source in order to make all the self-driving cars compatible with theirs. Compatibility issues among cars from different manufacturers might prevent cars from communicating, for example..

On the other hand, misunderstandings are to be expected from non-self-driving cars and thus create traffic issues on roads. It is also possible that other brands will use this technology, improve it and make Google cars out of date.

What the future autonomous cars will look like?

As of now, no dominant self-driving car design has emerged. It is still at the first phase of Utter back and Abernathy model: the fluid phase. Products provide new features and functionalities but production costs are still very high. Because of these costs, only a few companies are developing similar products.

Tesla and Audi’s self-driving cars have a steering wheel and are auto-piloted. Being autonomous, the Google car concept totally removes steering wheels and pedals from the interior, allowing more free space to passengers and a new interior design. However, the shift in design from the traditional car to the self-driving car will probably be a slow process. The car has to be accepted by a large part of the population, which means that the car must inspire confidence to passengers. Thus, we can imagine that the dominant design will tend towards a traditional looking exterior and a living room interior as Mercedes experiment in one of its latest concept (picture below)[12].

640_mercedes-f015-2

Vans already had some success selling cars with an interior design looking like a living room. The concept was based on removable and flexible seats fixed on rails allowing passengers to create space depending on their moods.

To resume we can guess that the dominant design will probably be closer to the Tesla one than to the Google car.

 So how are autonomous cars made?

The first step of the value chain is the inbound logistics, except the raw materials, sensors and software are the specificity of the product. The car is not only composed of sensors and software but also cameras, GPS, laser range founder, radars and wheel encoders. We think the car is made like traditional cars, but with an additional part in engineering. The engineering part is the most important for Google, who need high-tech specialists to produce the car.

Who is the leader in our technology?

As Google car is not on the market, Tesla is the current leader in the market. They have a lot of comparative advantages, Tesla cars’ owner can charge their electric car for free in specific stations. Furthermore, on those charging stations, called superchargers[13], everything is done for the client’s comfort: restaurants, Wi-Fi.

However Tesla cars remains normally drivable whereas Google car will be fully self-driving. Nevertheless regarding autonomous car market, supply chain and retail services, Tesla can be considered as the leader on autonomous car market today. People can wonder which of these two companies is leader on the self-driving car segment: Google is the leader technologically speaking because it is far more advanced. Google car is more disruptive by its philosophy than Tesla’s putting them on the cutting edge of the self-driving technology, and by the way leaders. On the market, Tesla can be considered as a ghost hunter. If Tesla managed to create its own software, they are still running after Google’s technology.

We can use Teece’s model[14] to explain why Google creates ways of locking competitors out of the market..

Google has complementary assets, they have a strong brand image, people trust them and they can support Google car’s marketing. However there is a limit for the company because Google doesn’t have big factories for car production. Moreover, building a car requires a lot of engineering resources. Actual car mass producers have better established infrastructures to produce cars.

Google also has a strong appropriability regime: they are able to prevent imitation thanks to patents and trade secrets. They created the most developed prototype of self-driving car and they have a patent for every details. Only few companies in the world have enough funds to develop such a complex project.

Since the beginning of the project, Google hired the best computing engineers to create their main asset: the car software. Google is known as a very comfortable company for its employees[15].  It is hard for its competitors to hire the engineers able to imitate Google’s technology advance in the self-driving vehicles project.

What will be determinant for Google Car adoption by the market?

Rogers’ has come up with 5 characteristics of innovation that explain different rates of market adoption

 

  1. Relative advantage: this factor applies to this technology because of the big advantages and differences compared to usual cars. People don’t need to know how to drive, it should be safer, more ecofriendly than conventional cars.

 

  1. Compatibility: People will use the current roads with future autonomous car so it is compatible with the existing structure. It is consistent with actual values and experiences because people will still use their car to go from one place to another. However a psychological barrier remains: People will no longer drive and will they accept to use those autonomous cars, leaving the control of the vehicle to sensors and software. It raises juridical and ethical issues in case of accidents: whose responsibility is involved?

 

  1. Complexity: Our society tends to want every new device to be easy to use. Just think of how quickly users set aside complicated devices. As a consequence, the Google car must be easy to use—and Google engineers are well aware of that. Consequently people may not know how the car actually works but will be able to use it easily.

 

  1. Trialability: Google car trialability does not exist yet and will probably be rather low when it launched on the market. First Google cars on the market will be rare and limited to certain areas such as main cities and few people will have the opportunity to try it. On the contrary everybody can test a smartphone app when unveiled. As it is a material product, its beginning on the market will be restricted, only car dealerships will offer the opportunity to test drive it. Instead of one might consider that Google will keep the control on its car and thus open a new type of dealership.

 

  1. Observability: One may consider that the Google car is quite visible. Actually only a few of them exists and are completing a lot of tests in a specific location where they are allowed to drive. As those cars are not yet available on market, the Google street view cars is a good reference because everybody is looking at them on streets. In terms of observability it is a good example of the reason why the Google car might spark the interest

 

All of these five factors can be applied to the car and as a result Google will managed to catch customer’s curiosity and attention very quickly and succeed at reaching an important market.

What might be Google’s strategy to enter the market?

First we decided that the segment in which our technology is applied is young and dynamic managers and engineers, kind with new tech and innovative environment. Those people are working and living in big cities. People working in banking, finance or very strict environment are not our target. The typical early adopters are, in our opinion, young engineers, start-up workers, adman. A lot of people who are living in town don’t have any driving license but may need one for business or even leisure. According to the actual estimated Google car’s price, these early adopters will be relatively wealthy. Those early adopters will create a community even a sub-cultural way of life that will influence the first early majority, this will allow Google to cross the chasm between early adopters and early majority.

As we already discussed earlier in this article[16], another segment could emerge, in the freight transport industry. This technology can quickly be implemented in trucks, opening up another segment of customers.

The adoption in the professional sector will probably not be without difficulties in the professional sector. Workers union will most likely strike against the replacement of drivers by Artificial Intelligence, the shift has to be done softly and smartly.

 

[1] http://gpmetropole.fr/blog/un-destin-automobile-la-voiture-autonome/
[2] https://commons.wikimedia.org/wiki/File:DARPAGrandChallenge2005-StanleyAward.jpg
[3] https://sg.news.yahoo.com/driverless-cars-coming-singapore-not-040044913.html
[4] http://read-it.hu/category/technika/
[5] http://spectrum.ieee.org/automaton/robotics/artificial-intelligence/how-google-self-driving-car-works
[6] http://spectrum.ieee.org/automaton/robotics/artificial-intelligence/how-google-self-driving-car-works
[7] http://www.cdc.gov/injury/wisqars/overview/key_data.html
8 https://www.theguardian.com/technology/2016/mar/09/google-self-driving-car-crash-video-accident-bus
[9] Theory developed in his book, The Innovator’s Dilemma.  
[10] Features: 25 mph max speed, 2 seats, small trunk, autonomy…
[11] http://www.theverge.com/2016/4/7/11383392/self-driving-truck-platooning-europe
[12] http://www.welt.de/themen/continental/
[13] https://www.teslamotors.com/supercharger
[14] Theory developed in the article, Profiting from technological innovation
[15] Ranked 8th in Glassdoor’s 50 best places to work
[16] see “Is Google Car disruptive or not?”

 

An article written by:

Corentin Charneau, Charline Laroche, Jeanne Beuscart,
Alexis de Guillebon, Catherine Chourpa

 

Leave a Reply

Your email address will not be published. Required fields are marked *