The Road Ahead

The self-driving car has migrated from the pages of science fiction to industrial research and design laboratories in a surprisingly short period of time. Eric Payne looks in greater detail at how this nascent industry is advancing and what it might mean for the future of transportation and logistics.

Robots are much maligned in our popular culture. This probably stems from the origin of the word ‘robot’, which was first used by a Czechoslovakian playwright in the 1920s to express his anxiety about the dehumanising effects of industrialisation and mechanisation. In the East and in Japan in particular, things are different. Robots tend to be associated with Astro Boy, a manga series first published in the 1950s about a mechanical boy who is helper and friend to his master Dr Tenma. This goes some way towards explaining the wide scale adoption of robotic systems by Japanese factories in the 1980s and 1990s, which resulted in a dramatic increase in productivity, bringing about an economic boom.

These cultural sensitivities may be even more important as robotic systems come to play a more prominent role in our everyday lives. One of the most interesting exemplars of this phenomenon is the automated or autonomous vehicle. Companies such as BMW, Audi and Mercedes already incorporate computer-controlled systems into high-end cars as a form of ‘driver aid’ – collision detection and driverless parking. But how far are we from a world in which cars ‘drive themselves’?

Hazard avoidance

Computers play an important role in almost every technology we encounter, and cars are no exception. The latest Toyota Prius is estimated to contain some 100 million lines of code, which is already more than some jet fighters. Automotive manufacturers are in the process of developing and introducing a range of quite advanced driver assistance systems with the potential to transform modern day transportation. Many vehicles are already equipped with traffic sign recognition, collision avoidance systems and automatic parking assistance. GM Vice President of Global Research and Development Alan Taub, speaking at the Intelligent Transport Systems World Congress in October 2011, said: “The technologies we’re developing will provide an added convenience by partially or even completely taking over the driving duties… Future generation safety systems will eliminate the crash altogether by interceding on behalf of drivers before they’re even aware of a hazardous situation.”

The biggest question that remains to be answered about these systems is one of safety. Almost any coder will concede that all software has bugs, regardless of expense or sophistication. In fact, the larger and more complex a piece of software gets, the more cumbersome and ‘buggy’ it is likely to become. This can result in emergent effects, producing unexpected outcomes, which, when said software is controlling a vehicle with multiple occupants travelling at 70 mph on a busy motorway, a bug could be the difference between life and death. Indeed, complaints by Toyota Prius owners about ‘phantom’ acceleration and break problems widely reported in 2010 were most likely the result of software error.

Providing customers with the assurances that they need to relinquish control of the wheel to a machine is therefore essential and the only way one can foresee of achieving this is to test, test and then test again. To that end, in June 2011 the Nevada Legislature passed a law to authorise the use of autonomous vehicles, becoming the first state where driverless vehicles can be legally operated on public roads. In effect, Nevada has turned itself into a laboratory for autonomous vehicle research and one can probably expect results from said tests to be coming thick and fast now that a government authority has taken this step.

With even industrial stalwarts embracing autonomous vehicle technology, the automotive industry appears united in its direction of travel, but will it be bold enough to embrace the full implications of what its efforts have so far wrought? The answer is that the automotive industry may not be allowed to make that decision.

Enter the billionaires

The measured pursuit of a new automotive status quo that incorporates driver aids has been blown apart in recent years by the intrusion of a couple of very rich California geeks. Not content with disrupting the computing, media, telecommunications, mobile phone and publishing industries, what began as a ‘hobby’ for Google’s eccentric billionaire owners – Larry Page and Sergey Brin – has become an earnest endeavour to reduce road deaths by incorporating advanced artificial intelligence and automation technology into cars. In October 2010 Google published a blog titled ‘What we’re driving at’, which announced, in typically grandiose fashion: “According to the World Health Organization, more than 1.2 million lives are lost every year in road traffic accidents. We believe our technology has the potential to cut that number, perhaps by as much as half.”

Noble aims and global ambitions aside, Google’s engineers have probably accelerated full-scale commercialisation of autonomous vehicles by several decades, leveraging a skill set and a culture of innovation not found at any of the major automotive OEMs. The commercial pinnacle of its work to date was in December 2011 when Google patented a technology for switching a vehicle from a human-controlled mode into a state where the car itself takes charge of the wheel. From a technical perspective however, this is just the beginning. Over the past 18 months, Google’s self-driving vehicles – six Toyota Prius and an Audi TT – have logged over 140,000 miles travelled. The cars have driven around Google’s Mountain View campus, crossed the Golden Gate Bridge, navigated the Pacific Coast Highway, and even made it all the way around Lake Tahoe (almost) without incident. The cars’ record was tarnished slightly in August 2011 when an over-enthusiastic motorist touched bumpers with one of the vehicles at a set of traffic lights. Although Google has stated that the car was being driven manually at the time of the accident.

This singular incident notwithstanding, Google’s self-driving vehicles are among the most sophisticated autonomous robotic systems yet created and a fascinating insight into what a future where people no longer drive themselves might look like. The ‘heart of the system’ is a 360-degree range finder mounted on the car’s roof, which uses lasers to create 3D models of the surrounding environment that are then matched to high-resolution maps – collected using manually driven vehicles as part of the Google Street View project – in order to calculate routes and avoid obstacles. The automated cars use video cameras, radar, GPS and all manner of sensors to ‘see’ other traffic and navigate the road ahead – and this 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.

A turn for the better?

The range of potential applications for this technology is obviously very wide. One could easily imagine autonomous vehicles being used to provide bus, taxi or road freight services, and the fact that established automotive companies like BMW, Audi and Mercedes are already incorporating elements of this technology into their road cars is a sure sign that Google’s efforts at the forefront of artificial intelligence and robotics research are likely to have a commercial future. Consumers may be sceptical about, or even actively resistant to the idea of handing human agency over to a machine, but, given the utility and safety arguments, not to mention the economic case, self-driving or autonomous vehicle research is likely to continue to receive significant investment.