By MIKE WHITFIELD
MD at Nissan South Africa
Autonomous driving technology is developing so rapidly that Business Insider’s research has forecast 10-million such cars on UK roads by 2020.
However, the closer we get to completely driverless cars, the more critical it becomes for automakers to make sure they’re safe.
Autonomous cars can save lives. Not only should they reduce serious crashes (the UK’s Society of Motor Manufacturers and Traders says the number of crashes will drop by 25 000 a year by 2030) but will also make cars available to people previously unable to drive. (Which begs the question – what will happen to driving-ban punishments? – Ed)
WHAT ABOUT UNPREDICTABLE SITUATIONS?
As the tech advances so important questions about the complexity of having them on the road will continue to arise. For instance…
- How can drivers learn to trust an autonomous vehicle?
- How will the car communicate with its humans to warn of nearby vehicles?
- What action will it take having identified an objects, sign or other road infrastructure such as painted lanes?
- Can a driverless car handle an unpredictable situation?
A major question: what if there’s an unpredictable situation? Will the computer navigate safely through the scenario?
Current (2017) autonomous tech is not fully autonomous. Nissan’s ProPILOT, for instance, still requires a driver ready to take control.
Technology on sale in Japan since 2016 enables cars to drive autonomously in single-lane, including in heavy stop/go traffic. It’s the first time a combination of steering, acceleration and braking has been operated in fully automatic in heavy traffic to ease the driver’s workload.
(Not strictly true: Mercedes, for one, has had a such a system for some years. – Ed)
However, ultimate control and responsibility remain with the driver.
If the driver releases the steering-wheel a warning light will glow and an alarm sound then the system will deactivate.
The day is fast approaching, though, when completely driverless cars will be a reality. Then the question of who takes control in an emergency will need to be answered – particularly if an ethical decision is needed… should you swerve to avoid a pedestrian and so endanger your passengers?
It’s a major problem still to be solved.
SAM IS RIDING TO THE RESCUE!
Good news, however: Nissan’s ‘Seamless Autonomous Mobility’ (SAM) system can navigate a surprise situation such as a collision or road works. SAM will help to realise a future in which autonomous cars can operate safely and smoothly.
How does SAM work…?
It’s smart enough to know when not to navigate a potentially dangerous situation by itself. Let’s say there’s a crash scene where police are using hand signals contrary to the road rules?
LEARNING THROUGH EXPERIENCE
SAM will halt the car and request advice from the “mobility manager” – the human in the car. (Tough if you’d hoped your autocar will take the kids to school and bring itself home again. – Ed) The mobility manager will paint a virtual lane for the car then, clear of the situation, it will resume control.
SAM can also learn from experience so, as autonomous technology improves, vehicles will require less human assistance; he coul aaccelerate by decades the introduction of autonomous vehicles.
Watch this space! – Mike Whitfield