Self Driving Cars
“An autonomous car, also known as a personal automated vehicle, or informally as a robotic car or self driving car, is a motor vehicle capable of automated driving and navigating entirely without direct human input. As defined here, automated vehicles exist mostly as prototypes as of 2013, but they are likely to be commercialized in the near future” – Wikipedia
The earliest representation of a driverless car was made by Norman Bel Geddes in his Futurama exhibit at the 1939 World Fair. Geddes' entry was sponsored by General Motors and was a prototype for an electric car which could be controlled by roadway circuits and radio. By late 2000s a number of automobile companies and researchers started to work on self driving vehicles. The major names working on the development of these autonomous vehicles include Google, Audi, Nissan, Continental Automotive Systems, Toyota, Bosch, and Oxford University.
It is estimated that by 2040 over 75% of the vehicles on US streets shall be autonomous or driverless. In August 2012, Google announced that its driverless cars had made it across 300,000 accident-free miles and weathered various traffic conditions.
Radar Sensor – Radar sensors in the cars are meant to prevent reverse collision by automatic detection and to provide adaptive cruise control. With the use of adaptive cruise control the vehicle’s speed is reduced in keeping with the proximity of other vehicles and objects. Throttle control, braking and transmission downshifting are used in combination for the control functions.
Roof Sensors – Sensors on the roof of the cars help in the collation of accurate relative distance data. These may be transmitted to a central navigation network for a coordinated driverless car control system in the years to come. The use of V2V (vehicle to vehicle) and V2I (vehicle to infrastructure) communications will grow in importance, experts believe.
Cameras – Electro optical and infrared cameras are fitted to the driverless cars. Stereo cameras impart the cars a three-dimensional detection ability making driverless navigation a safe prospect.
Software – Autonomous cars use techniques such as GPS and LIDAR to sense their surroundings and process thousands of data points each second. Apart from locating the vehicle’s positioning, these also help in interpreting movement and actions in the vicinity.
Steering wheels, brakes, and gas pedals could soon become a thing of the past. With the evolution of unmanned autonomous vehicles, all devices or tools used as human interaction interfaces with cars could soon disappear.
Car Accident Statistics
According to 2009 data, there have been over 5.5 million motor vehicle accidents in the US. Over 9.5 million vehicles have been involved in these accidents in which 33,808 people have lost their lives. In all about 2.2 million have been hurt and among these over 240,000 victims needed hospitalization. The country as a whole bore a cost of $450 billion for such motor vehicle accidents. In 2010, car crashes alone have been the cause of death for over 35,000 people. Worldwide over 1.27 million people die in car crashes each year.
It is estimated that 90 % of all car accidents are caused by human error. Drunken driving is one of the leading causes of car accidents. According to a 2011 Center for Disease Control and Prevention (CDC) survey, over 4 million adults in the US drink and drive each year. Self driving cars promise to minimize accidents by eliminating the need for human control. They have better reflexes and awareness, say experts. According to a study the use of autonomous brakes will reduce the risks of car crashes by 14%. Over 81% of all road traffic accidents can be avoided by the efficient use of vehicular communication systems such as networks that share traffic information with all vehicles. Self driving cars could totally eliminate the risks of any accidents caused by tired, sleepy, and distracted drivers. Impaired drivers are likely to find a dependable solution. Accidents may not be completely avoided but will be greatly reduced by self driving cars, researchers say.
A Look At The Environment
According to 2011 data, the United States consumed over 134 billion gallons of gasoline. This comes to about a daily average of 367.08 million gallons. Highway congestions costs the country $87 billion and 2.8 billion gallons in wasted fuel. About 22% of all Carbon dioxide emissions across the world come from trucks and cars. Self driving car exponents believe that they may have a possible environment-friendly solution. Over 75% of the emissions in cars are caused due to the weight of the vehicles. An average driverless car could be as light as 1000 lbs vis-à-vis a 4000 lbs regular car. Driverless cars are likely to bring down fuel emissions by 56% and hence could be very environment friendly. According to some studies, driverless cars could lead to savings of 1.9 billion gallons fuel in the US and over $101 billion saving on fuel costs and lost productivity. Car utilization is likely to go up between 10% and 75%. This could mean up to 90% less cars on the streets and a reduction of 80% in the cost per trip-mile.
Google Driverless Cars
Google driverless car is a Google X Lab project that has developed a technology for self driving cars. Led by director of the Stanford Artificial Intelligence Laboratory and co-inventor of Google Street View, Sebastian Thrun, the project has gained much fame across the US. Other engineers working on the Google driverless car project include Chris Urmson, Mike Montemerlo, and Anthony Levandowski.
“Google says it does not want to make cars, but instead work with suppliers and automakers to bring its technology to the marketplace. The company sees the project as an outgrowth of its core work in software and data management, and talks about reimagining people’s relationship with their automobiles.” – NY Times, October 2012
Google believes that the market for autonomous cars in the US could be worth $2 trillion. Google, however, has faced much criticism too. While in Europe, driverless cars are being researched by automobile giants such as Volvo, Nissan, and BMW, critics question why an Internet technology company is keen on monopolizing the research in the US.
In his Forbes article "Why The Idea of a Google Driverless Car is Nonsense", Haydn Shaughnessy says "The tech press in the USA typically refers to the driverless car as ‘the Google driverless car’ and talks this up as an opportunity for the search giant to move beyond its current business model. I don’t think so."
Beyond The US
In the US, the three states of Nevada, California, and Florida have passed laws approving the operation of autonomous cars. In June 2011, the state of Nevada passed a legislation which went into effect on Mar 1, 2012. The first license for a self driving car was issued by the Nevada Department of Motor Vehicles in May 2012. Other states that have bills pending legislature are District of Columbia, Hawaii, New Jersey, and Oklahoma.
Apart from the US, autonomous cars have been tested on European highways as well. In May 2012 Volvo ran a fleet of driverless cars through the highways of Spain. The automobile maker promises to launch cars to pilot this project codenamed SARTRE (Safe Road Trains for the Environment) by 2014. Nissan, the Asian automobile maker, has promised to launch its own technology by 2020.
Asian markets, however, seem to be of concern to autonomous vehicle technology developers. Concerns are that a number of countries may not allow Google to collect data to be incorporated into maps without which the technology may fail. Countries such as China and North Korea may be closed to such technology while other markets such as India may not have the requisite infrastructure to start with.
Weighing The Odds
In his article called “Here are some of the problems Google is having with its self driving cars”, published in the Business Insider in March 2013, Henry Blodget lists three challenges faced by the technology –
Difficulty in understanding manual or hand signals when traffic police direct the traffic
Hitches in navigation when faced with changes that are not on the map
Difficulty in detecting lane markers when it snows
Adam Ozimek’s article in the Forbes Magazine says “One important fact about self driving cars to remember is that right now they are being developed to work in a world that is not customized for self driving cars. Lane markers aren’t designed to be especially detectable by computers. Construction and road changes occur without concern as to whether the changes are registered on maps. Traffic cops are concerned that humans can understand their hand signals, but not computers”
According to Google, however, the self driving cars will reduce road traffic accidents by over 90%. There shall be 30,000 fewer deaths and 2 million fewer injuries by 2020. Over $400 billion shall be saved in accident-related expenses since 4.95 million less accidents are likely to occur.
Auto insurance companies collect over $200 billion in personal auto premiums. With the proliferation of driverless cars and reduction in accidents, the need for auto insurance may reduce considerably.