Autonomy Has Arrived, But Is It Ready?
A near-unanimous chorus of automobile manufacturers, robotics engineers, urban planners, economists and various experts predict that fully self-driving motor vehicles are looming on the horizon.
Telematics devices and systems – GPS navigation, tracking and fleet management – have pointed the way toward vehicle automation, along with older technologies such as cruise control. More development is sure to follow.
There are different opinions as to exactly when the first autonomous models will be ready for sale; some automotive executives claim that self-piloted cars will be rolling off assembly lines as soon as 2019 or 2020.
Others say that working out all the software bugs, proving the vehicles’ reliability and safety and – this is a major hurdle to overcome – untangling the yet-to-be-determined legislation and liability issues related to autonomous driving (such as: in an accident, who is at fault — owner or manufacturer?) will push the road date for these vehicles decades into the future.
However, nearly all agree that self-driving vehicles are inevitable.
Telematics and the autonomous vehicles
The advantages are too great to ignore. More than 90 percent of motor vehicle accidents are the result of human error, and placing the task of driving into robot hands – actually, a technology that replaces hands – will save countless lives. A number of medical patients today willingly undergo remote-control or automated surgery, and if people can put their faith in machines to carry out a procedure this personal, allowing some form of artificial intelligence to drive an automobile looks a lot less threatening.
In addition to safety, there are other gains. When cars operate themselves, they will communicate with each other and with stoplights. The resulting coordination of movement can all but eliminate traffic jams. It will almost certainly decrease commuting hours and fuel consumption.
That is, if cars and trucks are still burning petroleum by this time.
Nevertheless, this imagined utopia, or autopia, can’t happen all at once. When autonomous motor vehicles finally go on sale, not everybody will rush out and buy one even if they wish to do so. According to United States Bureau of Transportation statistics, the average age of passenger cars on American roads in 2016 was 11.6 years. This number indicates that plenty of automobiles 20 years old or older are still roaming the highways. For the typical U.S. citizen, an automobile either is the most expensive or second-most expensive investment made, and one that’s not replaced frequently.
It is also an unrealistic assumption that when these new cars are available, everybody will want to join the revolution. A 2017 survey reported in Forbes found that 78 percent of U.S. drivers feel anxiety at the prospect of riding in an autonomous vehicle.
Not all consumer reluctance is motivated by safety concerns. We can expect a certain amount of resistance from hobbyists, restorers, collectors of various relics and those who find pleasure in driving. (Yes, commuters – there are people who enjoy driving a car.)
These holdouts are not delighted by the prospect of automating another technical skill. The first practical automatic transmissions became available in passenger vehicles in the late 1930s. Today, several manufacturers still successfully sell cars that are equipped with a gearshift lever and a clutch pedal. This includes both economy makes and high-performance models.
For these reasons of cost and consumer preference, motor vehicle autonomy will be a gradual progression rather than occur all at once. For several years, perhaps decades, autonomous cars and trucks will travel the highways alongside human drivers. That human-machine interface poses all sorts of challenges.
A scholarly paper published first published in 2007 and titled Sharing the Road: Autonomous Vehicles Meet Human Drivers, describes possible control systems for managing roadways where human-directed and automated vehicles meet. One of these ideas is that traffic signals will operate on what the authors name the FCFS model, for “first come, first served.” Traffic lights would remain red in all directions until motor vehicles approach, turning green for the first one to get close to it.
Autonomous vehicles and human error
That appears to be a perfectly logical solution – but it fails to take into account the way that quite a few drivers behave. In a world where motorists increase speed toward a yellow light in order to squeeze through before it turns red, what will happen when an autonomous vehicle and one controlled by a person reach this signal?
Some have already guessed. The human driver knows that autonomous technology is programmed to stop its vehicle immediately if it detects an obstacle in its path. Armed with this knowledge, drivers and pedestrians may ignore their own red light and deliberately cut in front of the autonomous automobile – counting on it to slam on its brakes and allow the red-light runner to pass through the intersection first.
A popular prediction regarding autonomous vehicles is that their occupants will enjoy reduced tension and frustration. That seems unlikely in the situation described here.
We are witnessing the start of these encounters. In November, a minor collision took place in Las Vegas between a driverless shuttle and a conventional delivery truck. The truck backed up in the path of the shuttle, which promptly stopped. The truck didn’t — and ran into the shuttle vehicle. No serious damage was done but as this event illustrates, even fully self-driving technology can’t completely compensate for human error.