There have been several announcements concerning the connected car this week, but two from Google and Volvo are the most interesting. Both companies have demonstrated the abilities of their autonomous cars, which could massively reduce congestion by removing human error and inattention from the rush hour commute.
If a queue of traffic was to begin moving as soon as the traffic light turned green, congestion in cities could be drastically reduced. As it stands, drivers begin to pull away when the vehicles in front of them in a queue begin to move, meaning that a light can switch from red to green and back to red before the car at the back of the queue has begun to move. This is massively inefficient but it is something that could change imminently.
Google’s self-driving autonomous car has made headlines before. In 2012, a dozen Google cars completed 300,000 miles of driving without being involved in an accident ‘ guided by cameras, lasers, radar and GPS. The only incident occurred when a car was navigating a car park under human control, resulting in a minor collision.
The exemplary safety of these vehicles comes from the omni-directional and continuous scanning of the road and surrounding environment. Software on board is analyzing every object and change to ensure that the car travels safely. Google has a series of impressive videos on its blog that pair the on-board video with footage of the computer system at work ‘ tracking the objects around it and responding accordingly.
The Google cars have now logged 700,000 miles on public roads, using software that has been updated to predict likely occurrences while travelling. Google said that ‘what looks chaotic and random on a city street to the human eye is actually fairly predictable to a computer. We’ve built software models of what to expect, from the likely (a car stopping at a red light) to the unlikely (blowing through it) but we still have lots of problems to solve before we tackle another town.’
Google says it has made great progress, as ‘thousands of situations on city streets that would have stumped us two years ago can now be navigated autonomously. We’re growing more optimistic that we’re heading towards an achievable goal ‘ a vehicle that operates fully without human intervention.’
Also this week, Volvo has also been testing its autonomous cars in Gothenburg, with 100 vehicles deployed as the Drive Me project. The technology, called Autopilot, can currently handle lane following, speed adaption and merging traffic autonomously. It is being trialed by regular Volvo customers, not company engineers and technicians.
Swedish authorities approved 50 kilometers of roads that could be used for testing, which include motorway conditions and frequent queues. The feature is currently being envisioned as a mode that can be activated by the driver, and deactivated when needed.
Such a system is essentially an evolution of cruise-control, but with radar telling the car what speed to travel at rather than a manual input button on the dashboard. This implementation will almost certainly appear in Volvo cars at some point, with rival versions almost guaranteed, but a fully autonomous vehicle is a much more enticing prospect ‘ it removes the human fallibility and idiocy from behind the wheel completely.
Safety records that are hampered by human error (speeding, intoxication, distractions) could be eradicated if it was not a person doing the driving. Fuel efficiency could be boosted too, cutting carbon emissions, by programming the cars to drive in the most fuel-efficient manner possible.
This would be good news for cities, which are often plagued by high pollution levels due to the number of vehicles stuck in traffic with their engines running. Autonomous cars with intelligent management could help improve air quality in city centers while reducing congestion and travel times.
This functionality could be improved by adding a system that lets a car check a database for the most convenient parking space. A driver could arrive at their destination, by driving themselves or travelling autonomously, before leaving the car in ‘valet mode,’ whereupon the car would drive off to find somewhere to park. A simple app could recall the car when needed (‘Pick me up from work at 17:00’).
There are already companies working on systems like this. WorldSensing has Fastprk, Urbiotica has U-Spot, and Kiunsys has developed the rather practical if uninspiringly named Parking Spot Sensor. All three systems operate in much the same way ‘ monitoring and plotting the available parking spaces in an area and conveying that information to those who need it.
Another application for autonomous cars is in public transport. A fleet of vehicles could be used much like the current system of taxi cabs, most likely configured by a smartphone app that lets users hail a car and pay for their trip.
An additional implementation would be long distance freight haulage. The potential for disaster increases exponentially when heavier vehicles are involved in a crash, so there are obvious and compelling arguments for improving safety among HGVs by converting them to drive autonomously.
The issue of blind spots and tiredness due to overwork would be negated by a self-driving truck. Similarly, congestion could be avoided by scheduling the traffic to run at night, when the roads are quieter.
There are of course many obstacles to the introduction of these vehicles. The first is the cost, which means that it will take many years for self-driving vehicles to comprise the majority of cars on the road. This is simply because people upgrade their cars over years and sometimes decades, not months.
But the age of private ownership of cars could come to end if the public transit model came to pass. Consumers would not have to pay for the maintenance of their vehicle, or the insurance, or the cost to buy it, or the taxes paid on it, or the cost of acquiring a license for it. In fact, a self-driving taxi could save the average consumer a lot of money when paired with subways and buses, while still providing the environmental benefits of automation.
There is also an economic argument to put forward, as self-driving vehicles could put entire industries of workers out of their jobs. Responsibility in case of an accident is a complicated issue, with liability needing to be established in a court before insurance companies are comfortable backing the driverless vehicles.
Current tests keep a person behind the wheel at all times so that insurance policies are still valid, even though the only reported accident involving an autonomous car occurred when the Google tester was in control, but keeping a person at the wheel at all times would undermine the design of an autonomous vehicle.
The potential for driverless vehicles is undoubted, but social acceptance to the idea might mean it is still a long way off. The greatest obstacle is not the technology itself, but the motoring society in which it must take root. Winning the public over might be a bigger task than piloting thousands of pounds of metal and plastic along a road using microwaves without killing anyone.