The Forbidden Roads

Robotic Cars and the End of Bicycles and Pedestrians

Have you seen any horses on a paved road lately?

"Self-driving" automated cars are still rare in 2016 as I write this, experimental and expensive, the domain of technology companies demonstrating their future technological competence to stockholders. The stories about robot superiority over human drivers are generated by public relations departments, propagated by broadcast media, and accepted as inevitable by the non-technical public. Robots outperforming all humans in all mixed human-robot environments has not been demonstrated, but robots do outperform drunks and idiots in the restricted environments (like controlled-access freeways) where the robots rack up the most mileage.

Assume that robotic cars become good enough to navigate most roads with a tolerable number of accidents, and that the spinmeisters successfully sell their successes and obscure their disadvantages. When the spinmeisters control the world's principal websearch engine and most of the smartphones that access it, this shaping of public opinion will be inescapable. What next?

If robotic cars are networked, they will rarely collide with each other. Few of their collisions will be with human driven vehicles, because other automobiles are big easy to detect. When robot-human vehicle accidents do happen, it will mostly be with impaired humans, because impairment is the cause of almost all human-human accidents (stupidity and bad judgement are impairments).

The first consumer robotic cars will be expensive - totalling a robotic car in an accident will be very costly. The insurers of the robotic cars will work with the robotic car manufacturers to create a legal environment where the fault is assigned to the human driver and the other insurance company, using public relations, lobbying, out-of-court settlements and expensive attorneys to shape the outcome of every case - regardless of the facts of the case. The expense of robotic cars will be offset by the increased expense of insurance for the remaining human-driven vehicles, whose skyrocketing insurance premiums will pay the costs of the robotic transition. The remaining non-robotic cars will be hauled (by robotic trucks) to the recyclers. Vehicles will cost more - electronics is not free - and used non-robotic cars will disappear from the road.

This will lead to "optimizations". For example, why space cars "safely", when a platoon of robot cars can travel bumper to bumper, accelerating and braking in unison, saving energy by travelling in each other's slipstream? Why stop at intersections, when the robot cars can regulate their speed with high precision and pass through gaps in cross traffic, with centimeters of margin? Why use primary roads, when shortcuts on secondary roads through unfamiliar neighborhoods can shave minutes off a trip? With braking and acceleration minimized, roads will be used more densely, so road construction and maintenance costs can be reduced. Lanes can be narrowed, or robots can smoothly veer sideways to avoid well mapped potholes (or other fixed obstacles) hundreds of meters ahead.

For that matter, if the robots always avoid each other, who needs the expense, weight, and unnecessary extra vehicle length of bumpers?

In this new non-stopping, collision-avoiding road network of smooth, automated, reduced friction travel, there will inevitably be incentives to minimize travel time, and eliminate delays. In the worst case, this will create political pressure for higher speed limits on city streets, increasing energy consumption, mechanical failure, and threats to non-vehicles. More likely, it will result in legislation that removes "obstacles" from the robot paths.

If travel by motorized vehicle becomes easier and faster, those who are willing to pay will do a lot more of it, and consume many more road miles while they do other things in their isolated travelling bubbles. Ordinary white collar workers will probably be expected to work while they travel, earning the extra income they need for energy/fuel, more expensive vehicles, and paying for copyrighted entertainment (a frequent justification for robotic cars).

When something becomes cheaper, people spend more on in total on it. Robot cars may be more expensive to purchase, but the effort to use them will be minimal, so the incremental cost of using an already-purchased robot car will seem small. People who can borrow enough money to "afford it" will spend their texting and web surfing and television time in their vehicles, perhaps even sleeping in them. Some may never leave their single-occupant robot cars.

Non-robots in a robotic world

Bicycles, pedestrians, and animals live on islands in a sea of pavement. They must cross the automobile-dominated pavement many times to reach other islands. Biological brains are not networked, slow and erratic and unpredictable. Brains may not operate the motor vehicles of the future, but they will always challenge them.

I live between two four-lane state highways, sandwiched between freeways. It impossible travel more than a mile from home, on foot or by bicycle, without crossing those highways. The crosswalks are half a mile apart. Crossing the road to a bus stop on the other side can take 10 minutes, including the two minute light cycles and stopping to cross side streets. While some drivers will (unpredictably) stop for a pedestrian or cyclist crossing between crosswalks, it is rare that all four lanes of traffic will. More often, speeding drivers will go around the courteous ones, and many pedestrians have been killed on both highways this way.

If pedestrians and cyclists carried GPS trackers, the self-driving cars could avoid them. But how do robots determine whether the pedestrians plan to cross, or have merely stopped to look at something? Can a robot detect what a human plans to do? Even humans with Asperger's have trouble with that. Like Google's robot car prototypes, the mind-blind have problems with tailgating and rear-end accidents, another situation where the intent of humans is difficult to determine. Add children, pets, and wild animals to the "pedestrian" traffic, and the problem may become impossible to solve without fencing off the roads entirely. For the convenience of robot operation, and the unbounded arrogance and laziness of their owners, the minority of car-free humans will be imprisoned behind chain-link fences.

Although my house is in unincorporated Washington County, Oregon, the highways (and the businesses along them) have been annexed by the tax- and speeding-ticket-hungry city of Beaverton. Decisions regarding through-routes will be made to optimize this revenue. Beaverton was, for a time, a "bronze" bicycle city, the lowest tier on the scale, but no longer aspires to even that lowest tier. Cyclists are "pedestrians" who must lift their vehicles onto the sidewalk to push the crosswalk button. Brief segments of bicycle lane are disconnected, interrupted by car sales lots that extend almost to the edge of the traffic lanes. Chances are, cyclists will be banned in many places, an impediment to the robots that cannot detect them.

"Roads" were originally for animals - ancient game trails coopted by people and their domesticated animals. The roads were widened and straightened for wheeled vehicles, then engines replaced the animals (horses and cyclists) powering those vehicles. Soon, robots will steer those vehicles. Someday, humans may be "unnecessary".

The end of an essay is where a prescription, a cure, should be offered. I have none. Developed-world cities will become 24x7 freeways, carrying increasingly unhealthy people from infancy to death by obesity. The far future may belong to cultures too poor to participate in the developed world's imprisonment and self-destruction.

EndOfPeds (last edited 2016-06-07 16:22:49 by KeithLofstrom)