We have all heard the debate over self-driving cars. Pundits, lobbyists, experts and lawmakers all weigh in with their perspective, or paid point of view, on an issue that will affect all of us. Aside from self-driving cars being an engaging problem to solve, not to mention a fascinating scientific advance, the number of avoidable human deaths every year from car accidents makes it a noble and desirable achievement. If it can be done. With faster computer chips and ubiquitous communications, we are seeing advances from military as well as commercial efforts. Each one chipping away at the challenge by approaching it from a different perspective.
Google’s self-driving car program is surely the most ambitious, and likely the most visible given the media attention it gets. Their objective is the toughest of all the self-driving challenges – to create a vehicle that will safely navigate open roads and traffic under any circumstances. It is a bold aim but may well be what ultimately demonstrates to the masses of drivers and pedestrians that a self-driving car can be safe and convenient. Initially, however, the small runabout with two seats and no steering wheel which many people have already seen in Google’s promotional videos may find it is first accepted within gated communities, campus environments or small towns The cars could be geo-fenced to stay within an identified area until they accumulate enough hours of incident-free driving whilst being commanded by regular folks.
Others have approached the challenge from another perspective. Volvo has already accumulated many hours testing long haul trucks in an application of self-driving vehicles. In Volvo’s case, a lead truck with a human driver runs along the highways with several other driverless slave trucks (or cars) following closely behind. The braking, accelerating and turning done by the lead or master truck is copied exactly, at just the right moment, by the slave trucks. This enables great cost savings not just from the elimination of additional human drivers but the computer-controlled slave trucks can follow much more closely behind one another in the slipstream that materially reduces drag. When the lead truck hits the brakes there is no hesitation by the slave trucks that all brake at the same time, with the same intensity. Warehouses are already situated close to highways which reduces the distance any road train like this would spend on regular roads, keeping all the computer controlled driving on the smooth curves and multiple lanes of highways with no traffic lights, cross roads or pedestrians.
We are already seeing high end luxury models of cars aimed at everyday drivers and families that incorporate the advanced technology of driver aids. A prelude to complete control?
Volvo has even applied some of this technology to self-parking cars. Where Ford and others offer park-assist that will parallel park a car for the driver, Volvo has painted a future where the driver will get out of their car at the curb in front of their house and the car will drive itself around to the garage either behind a house or in the underground parkade of a tower. The owner simply calls for the car to return to a designated spot when the vehicle is required again. Those parking garages will eventually ban human drivers who are far more likely to ding the sides of other cars getting in and out. The self-parked cars are stacked in there cheek by jowl because no one ever needs to open the doors.
We are already seeing high end luxury models of cars aimed at everyday drivers and families that incorporate the advanced technology of driver aids. A prelude to complete control? Perhaps. Many high end and even mid-level models have backup cameras, park distance sensors, adaptive cruise control and even warnings to keep the driver from drifting out of their lane. These technologies have led to the first incidence of a car automatically applying the brakes in an emergency situation. Lexus and other luxury car brands have been advertising a new system that will apply the brakes without the driver’s control or initiation when the car is in reverse and an obstacle is detected. That same feature is now being applied as “forward brake assist” in a situation where an accident is imminent so the computer applies the brakes to stop the car. This is a good example of the “technology creep” we are seeing in the latest car models. Starting with sensors and indicators that provide warnings when another vehicle is in your blind spot or an obstacle is in the car’s path, we are now seeing the car’s computer independently apply the brakes to avoid an obstacle in the rear path or an accident detected in front. Next will be control of the steering to keep the car from swerving out of its lane or off the road, or even active steering to swerve the car around an obstacle in an accident situation.
Once it is clear the insurance companies can save money and lives, then self-driving cars will become a requirement. It seems difficult to argue that computers will not easily exceed humans in reaction time, assessment of multiple threats and complete visibility around the outside of a vehicle at all times. Humans simply cannot compete.
This is all very interesting, and will progress to a point where these advanced safety features continue their technology creep into the newest models, and ultimately become as ubiquitous as seat belts or air bags. But it is the insurance companies that will become the real drivers of change (please forgive the pun). Once it is clear the insurance companies can save money and lives, then self-driving cars will become a requirement.
Make no mistake, driving will prove to be a “sticky” activity. Many will go down fighting before they allow themselves to be driven around by a computer on wheels. But insurance for all but self-driving cars will fast become prohibitively expensive. Driving a gasoline powered car by yourself will soon become something we do on a track, perhaps during a guys weekend.
It seems difficult to argue that computers will not easily exceed humans in reaction time, assessment of multiple threats and complete visibility around the outside of a vehicle at all times. Humans simply cannot compete. Then we just need to add communications between all cars to the mix. In the same way Volvo’s slave vehicles are told what is happening instantly by the lead truck, all cars will eventually speak to one another. When an accident occurs or one vehicle slows rapidly, all the nearby vehicles will respond appropriately. And on a highway, all the vehicles getting off at the next exit will be in the right lane while those going to the furthest destination are in the left lane. Adaptive cruise control can already eliminate traffic congestion and tailbacks on all highways, the only requirement is for every car to have it. Once more data can be broadcast and analyzed by all the surrounding vehicles they can all drive just feet apart from one another in fog or complete darkness.
It is not far away. As the many technologies get cheaper and more reliable they become included in more and more models. And as governments, or more likely insurance companies, require every car and driver to have and employ them, then my grandchildren will never learn to drive. They will instead simply call up an app on their pocket computer and request a car, either from a car sharing service or their own vehicle parked in the garage. The truth is, driving a car is far too dangerous to be left to humans to do.
One thought on “Why the insurance industry will drive adoption of self-driving cars”
I wouldn’t plan a career in the autobody business
Leave a Reply
You must be logged in to post a comment.