In Silicon Valley, Google experts are working on a scheme that could revolutionise the way we think about road travel forever. Andrew Hubbard investigates Google’s driverless car project.
For many years, successive governments have fought to reduce congestion on the UK’s roads and the exhaust emissions that the more than 34m registered vehicles create. Schemes from car-pool initiatives to congestion charges have made some impact in recent years, but still the challenge remains.
Over the Atlantic, internet search giant Google is working on a potential solution to this age-old dilemma through its driverless car project.
As the name suggests, the project aims to develop the technology that would allow human beings to travel in cars without the need to control them.
Google chiefs say that the self-driving cars will transform car sharing, significantly reducing car usage, and help to create the new “highway trains of tomorrow”. These highway trains would cut energy consumption while also increasing the number of people that can be transported on major road networks.
The project is being led by engineer Sebastian Thrun, vice president of Google and co-inventor of Google Street View. Thrun's team, based in Stanford, created a robotic vehicle which scooped the 2005 Defense Advanced Research Projects Agency (DARPA) Grand Challenge and its $2m prize from the United States Department of Defense.
Since 2010, Google’s team of engineers has worked to develop a fleet of autonomous cars, which today comprises Audi TT, Lexus RX450h SUV and Toyota Prius models.
The firm has ensured the right team is at the forefront of the project, placing Anthony Levandowski as Google's driverless car project manager. Levandowski is the man behind the world’s first driverless motorbike and also the creator of the modified Prius that delivered pizza without a person inside.
The project took a giant leap towards success in June 2011 when the US state of Nevada passed a law concerning the operation of driverless cars.
The company had been lobbying for driverless car laws for some time, but with the passing of this statewide regulation, Google was able to finally test its vehicles in real-world scenarios. The law went into effect in March, with the state granting the first license for a self-driven car in May, for the company’s modified Prius.
In August, Google confirmed that it had already successfully completed more than 480,000 accident-free autonomous-driving miles. The main goal of the project is to fundamentally change car use to help prevent traffic accidents, free up people’s time and reduce carbon emissions.
Any injuries or fatalities caused by the cars during the development phase would have an irreversible and catastrophic effect on the project’s future. It’s not suprising then that safety has been Google’s primary focus. To ensure that vehicles are 100 per cent safe the cars are never unmanned. Two specialists are in the vehicle at all times throughout testing: a trained safety driver behind the wheel to intervene should they need to, and specially trained software operator.
Before the team proceed with a test, a drive in a conventional car is undertaken to map the route and road conditions. By mapping features like lane markers and traffic signs, the software in the car becomes familiar with the characteristics of the environment in advance.
This cautious approach is crucial and is paying dividends. The first – and one of the only – accidents a Google driverless car has been involved with was in August 2011, when a human-controlled driverless vehicle crashed near Google headquarters in Mountain View, California. Google has stated that the car was being driven manually at the time of the accident.
According to the World Health Organization, more than 1.2m people around the world die every year in road crashes, and as many as 50m more are injured. Google believes that its technology has the potential to cut these numbers by as much as half, as the vehicles using it have the ability to process more information faster than a human driver. They do this by utilising video cameras, radar sensors and a laser range finder to see other traffic, and also by using detailed maps – collected using manually driven vehicles – to navigate the roads. It is all made possible by Google’s data centres, which can process the enormous amounts of information gathered by the cars when mapping their terrain.
The concept is that, in the future, driverless cars will be able to communicate with one another, allowing them to negotiate lane changes and overtaking.
The Productive Project
The Office for National Statistics has reported that the average worker in the UK spends nearly 200 hours a year commuting to and from work. Google says that not only will this project allow the passenger to utilise their commute time, but it will actually reduce the amount of time spent on the road as it eases congestion.
Where as traffic usually slows as the volume of cars on a road increases, the driverless cars could, in theory, drive at a much closer proximity and at higher speeds.
It would also open the road up to users that may not have been able to access it beforehand. The driverless car would grant mobility to drivers who feel uncomfortable getting behind the wheel at night, to those with poor eyesight and to people with paralysis, epilepsy, broken limbs or other disabilities.
The engineering team has saved much time in developing the project as far along as it has, something many are putting down to Google’s use of collaboration and technology licencing. Google knows where its strengths are, and by deciding not to go down the route of developing driverless vehicles from scratch – and instead modifying existing vehicles with technology – has saved years in terms of research and development.
It is a model that the company is relying on to take the project forward. According to the Wall Street Journal, Levandowski told the US Society of Automotive Engineers that the company was looking for partners to help it get the cars on the road within the next decade. He was quoted as saying: “The technology could be ready well before 2022. If not, shame on us as engineers.”
Despite speaking with a number of automotive firms, Google is still undecided on how it will make its technology available to consumers.
The company has, however, admitted that there are still several factors that need to be addressed, particularly with respect to who would be held accountable in the event of an accident. At present, the law in Nevada requires any driverless car to have a vehicle operator who shall be deemed the driver of the car, even if they are not physically present in the vehicle.
The law also stipulates that the named operator would be responsible for any laws that were broken by the car.
Despite all of the progress in the US, transport minister Justine Greening has recently gone on record saying that The Department for Transport had not begun working out how to license driverless cars in the UK.
Sergey Brin and Larry Page founded Google in September 1998 with one aim: to help solve the world’s problems by using technology. It seems that the work of Sebastian Thrun and his team on the driverless car project, Google may succeed once again.
A brief history of driverless vehicles
1977 - Tsukuba Mechanical Engineering Lab in Japan creates the first autonomous, intelligent vehicle.
1995 - One of Ernst Dickmanns’ robot cars takes a 1,000-mile trip from Munich to Copenhagen and back.
2007 - Carnegie Mellon’s Tartan Racing Team wins the DARPA Grand Challengew, after its driverless vehicle completes the 60-mile course.
2010 - Google driverless cars travel 140, 000 miles on California roads - including city streets.
2012 - Volvo completes a 125 mile journey using wireless technology to 'mimic' a lead vehicle.
2020 - the year by which GM's Alan Taub predicts autonomous systems will be standard equipment.
How it works
Position estimator - A sensor mounted on the rear wheel measure small movements made by the car and helps to accurately locate its position on the map.
Radar sensors - Four standard automotive radar sensors, three in the front and one in the rear, help determine the postion of distant objects.
LIDAR (light detection and ranging) - A rotating sensor on the roof scans more than 200 metres in all directions to generate a precise three-dimensional map of the cars surroundings.
Video Camera - A camera mounted near the rear-view mirror detects traffic lights and helps the car's onboard computers recognise moving obstacles like pedestrians and cyclists.
Stay in touch with the latest developments in your profession by subscribing to Project magazine, "The voice of Project management".
Non-members can subscribe for as little as £55 per year (10 copies), APM members automatically receive the magazine as part of their membership:
To order yours now, click "Subscribe" below:
See all posts