Can and should cars drive themselves? The world is beginning to come to terms with the idea of the autonomous automobile as improving sensors allow vehicles to "see" the road and increasing computing power allows them to make decisions.
Google's spin doctors would have us believe that it invented the idea (it didn't), but it's ironic that the search engine giant uses a Toyota Prius (along with a crew of ex-motor industry engineers and scientists) to conduct its research. Toyota, like the rest of the motor industry, has been working on the technology for decades.
Autonomous driving has long been a goal of car makers and legislators, with experiments going back to wire-guided cars in the Twenties. Fully autonomous cars, however, had to wait for the technology to catch up. There were early experiments in Japan in the Seventies, in Germany with Ernst Dickmanns of Munich University and VaMP. one of the first fully autonomous cars, the EU's Prometheus Project and the US Defense Advanced Research Projects Agency's (Darpa) Grand Challenge for driverless vehicles.
But if Google has moved the subject back into the spotlight (and Moritaka Yoshida, Toyota's chief safety technology officer, admits Google's pronouncements "may have stimulated us") the car industry reawakening to the driverless car. At the Frankfurt show in September Mercedes-Benz demonstrated an autonomous driving S-class and Germany has just given permission for it (and driverless rivals such as those from Audi, VW and BMW) to be tested on public roads – with an engineer at the wheel in case things don't go quite as planned).
Toyota has introduced a basket of technologies that bring the possibility of autonomous driving that bit closer, including a 360-degree view parking assistance system, a pedestrian braking system, an advanced form of cruise control (using wireless communication between vehicles) and a more accurate form of automatic lane-keeping.
Yet the industry is still cautious about the idea of a fully driverless car and points to improved safety as the major reason for such developments. Like Volvo and others, Toyota wants to create a road network free from death and injury (the uncrashable car), but there are significant hurdles
"The key element is safety," says Didier Leroy, chief executive of Toyota Europe. "We need to keep the driving pleasure for the driver, but we also want to keep the driver safe. We have the technologies [to make a driverless car], but the problem is making it affordable – and legal responsibility is a big issue. This is not a game. You aren't playing on a PlayStation with a second or third life."
We've seen vehicle-to-vehicle (V2V) short-range wireless communications prototypes before and Toyota's V2V enhanced cruise control works by sending information about acceleration and braking between vehicles to smooth progress and prevent the shunting that is so often a feature of motorway travel. It also increases safety by forewarning of braking ahead, while also reducing fuel consumption.
In more advanced versions it could also increase average speeds and allow vehicles to be electronically linked in platoons, which is one potential first step to full autonomous driving. Toyota's system smooths braking and accelerating and in a demonstration it appeared to work pretty well, but as development engineer Shota Fujii admitted, its widespread adoption depends on other vehicles having the same technology. And there's the rub.
Toyota's system works on the 700Mhz band and the company is attempting to persuade the world to adopt that frequency. At the same time, however, Toyota is one of eight car makers working in the CAMP research consortium which is looking into similar technology using Dedicated Short Range Communications (DSRC) 5.9GHz bandwidth set aside by the US congress in 1999. Betamax versus VHS, anybody?
The lane trace control is an all-new version of the more familiar lane-keeping assistance systems, using high-performance cameras, millimetre-wave radar, and control systems to adjust steering, engine torque and even braking to maintain station within a lane. Our driver removed his hands from the wheel of the prototype for several seconds through quite acute motorway turns, so we're happy to report that it works.
Both systems are currently on test in Japan and Michigan, North America. There is no on-sale date, but Yoshida dropped a heavy hint that by 2020, when Japan hosts the Olympics, he would like autonomous driving cars to be in use. "We will have a new age transportation system for then," he said.
Toyota's innovations are targeting pedestrians and older folk. It cites Japanese traffic accident data showing that, since 2011, pedestrian fatalities have overtaken those in passenger cars and are now, at 37 per cent, the most common form of fatality on Japan's roads. They also show that older people are disproportionately involved all accidents, either as pedestrians or drivers. In 2011 2,264 under-65s were killed on Japanese roads and 2,147 over-65s. Police data shows that older drivers have also been steadily causing more accidents on Japanese roads. There are also some slightly scary studies from the Japanese Society of Automotive Engineers showing that, compared with the average 30-year-old, a 60-year-old has significantly worse sight (particularly in low light), slightly less memory recall and powers of reasoning, and much slower reactions.
Toyota's answer to the pedestrian problem is to upgrade its existing pre-collision avoidance system (PCS) to include not just visual and audible warnings of a potential collision, and automatic braking if they are ignored, but also an automatic steering function to avoid a pedestrian while staying in the marked traffic lane. It also works for cyclists.
The sequence starts four seconds before potential impact and at up to 43mph. It uses an advanced Light Detection and Ranging (Lidar) technology, although engineers admit that the price and take-up of such systems is highly dependent on the cost of the sensors, and that subsequent systems might use simpler cameras as used by Volvo.
The basic PCS was introduced in the Lexus LS last year as a £2,000 option that will be gradually introduced across the entire Toyota range. The automatic steering component will be introduced as a similarly priced option in 2015. In demonstrations with a Prius on a test track, it worked adequately, although the conditions were highly artificial – and if you are struggling to stop your car or avoid an impact with four seconds notice, you shouldn't really be on the road.
Visitors to the 2020 Tokyo Olympics could be transported in driverless robot taxis, if an experiment with supermarket shoppers goes according to plan.
The Japanese government is supporting a fleet of “robocabs” in a trial due to begin later this year. Operated by Robot Taxi Inc, the fleet of driverless cars will carry shoppers about two miles from their homes and back along main roads in the city of Fujisawa, 25 miles south of Tokyo.
As a precaution, though, “crew members” will ride with the robot driver, ready to seize the wheel if necessary.
Shinzo Abe’s government will back the project in a conscious challenge to Google, which is experimenting with robot vehicles on US roads. Robot Taxi, a subsidiary of the gaming company DeNA, hopes to run the service on a commercial basis by 2020.
The robot taxis are seen as a solution to the problem of Japan’s elderly drivers, whose failing sight and slower reflexes are the cause of an increasing number of accidents.