Uber Etc

More speedbumps on the road to autonomous cars.

Who cares about one measly little human life? In a world of 7.6 billion people with more born every minute, any single one of them is statistically insignificant.

But it’s not.  On March 18, a pedestrian was struck by an autonomous vehicle. The car had detected the individual, but efforts to stop were delayed; the woman, whose body tested positive for amphetamines and marijuana, was struck by the Uber car and killed.

There is blame enough for all involved, and we haven’t heard the end of this, since we do not live in the Soviet Union of the 1930s.  Regardless of mathematics, we do care about one measly little human life, and the American inclination is to do everything we can to keep technology safe to use.

As a result, far-reaching decisions will be based upon this latest incident in the conflict between the rush for technological solutions and the ever-present need for safety.  However, because of the vast sums of money involved, there’s a risk that they will be manipulated such that technology wins, the safety of the errant human be darned.

Actually, it's a wonder that it has taken so long for a death to occur from a self-driving car. Uber Technologies has been testing self-driving cars for their fleet of taxis since February 2015 - an absolute eternity in the technology sector. And Uber has compadres in this emerging technology: Cruise Automation (GM) and Waymo (Google) are working on self-driving vehicles as well.

Aww, how cute!
...until you're stuck in one.

There is some external product differentiation: Waymo and Cruise are doing their experimentation on vehicles of the sub-econobox persuasion, whereas Uber seems to be concentrating on larger vehicles like Volvo cars and trucks including semis.

After the Uber vehicle struck the pedestrian, Waymo has announced a curtailment of its development activities; Cruise Automation has been silent on the topic. But this accident was a first. The details seem to indicate that there were enough screw-ups for the Uber car company to weasel out of direct blame of the technology; eventually someone of course will have some kind of liability for the accident in our litigious society, but the guilt is not immediately obvious.

The Psychology of the Powerless

There is no question that automatic cars are in our future. There will be schemes of ownership versus pay-per-ride; who will buy a car versus who will use it as a new form of public transportation, but the die has been cast.

A better driver than your teenager?

Whether we ride in an empty vehicle with an electronic driver hidden away in the “cloud” as is being tested by Waymo, or one with a pleasant “driver” who is somewhat preoccupied with his own smartassyishness as in the movie Blade Runner is yet to be determined. We do not know yet the psychology of riding in a car with no apparent entity in charge; there is no knowing what the implications of that will be.

The psychological impacts of being powerless in the single most dangerous thing that most people do daily are yet to be known. Some, like your writer, will experience the loss of control as an extreme negative and develop an uneasy feeling in general. Most adults have been driving for so long that their psyches expect to guide the car and make decisions, and all those micro-corrections that are needed for control and safety. Our mind is somewhat perturbed sitting next to the wife as she drives and makes those tiny corrections that drivers make, and the wife is hardly an automaton much less an invisible one.

For others, though, especially new or bad drivers, the frustrations and critical decision-making inherent in a quick drive to the store will never be missed. That is probably a good thing.

The Battle of the Bureaucrats

All of this is a sea-change for society. We don’t quite know how to handle even the thought of a change this drastic. These matters are supposed to happen gradually so that our tiny little brains have time to wrap themselves around the concept of a new way of operating.

And perhaps we will find a way of slowing things down to the speed at which we can cope. Perhaps even the hallmark delays of bureaucracy will find a welcoming home. Our public servants are probably even now ginning up ways of slowing things down unbeknownst to us. The AFSCME union may actually help for once.

One side effect of bureaucracy may be to change transport and commute patters.  London has already set up a system of toll roads such that cars going into the heart of the city pay a very large premium over trips elsewhere. There are beltway roads which form the boundaries between sectors of the city, and transition from one sector to one closer to the center requires a toll to be paid. It’s British, so in many ways, it makes absolutely no sense, and we reserve the right to have misunderstood it.

The intent of the London traffic zoning and tolls is to discourage traffic in the center city. The tolls are £11.50 ($15.35) per day, and while this is not a crippling sum it is an amount to think about. The expectation is that each car coming in will pay, and most of these cars will be used by one person for one trip in and one out.

An obvious dodge would be to provide downtown-only self-driving vehicles for transportation within that district, that pay the toll once but tool around downtown all day transporting multiple passengers, like a super taxicab.  Whether the bureaucrats and tax mavens will tolerate this reduction in their toll take remains to be seen.

The future will attempt many solutions to this transportation problem, and may never get it right: – look at the many attempts to incorporate traffic roundabouts into otherwise sane planning.

In fact, as conservatives instinctively realize, the bugaboo is the fact of the whole discipline of Planners. These government employees (for the most part, there are some privately employed) are thoroughly steeped in the authoritarian methods of Trotsky and the Communist Party – namely, that they know what’s good for you and you can like it or lump it.

This writer has worked in architecture/engineering firms with staff planners, and these guys give new meaning to meddlesome. They think nothing of making access to pieces of land verboten by tying them up in bureaucratic wrangling for decades. Just think of the wonders they could work in slowing down access to certain primo locations in downtown areas.

Government is deeply involved in the mix of transportation/vehicle types/vehicle preeminence, and this sector of our lives will be in constant flux typified by high-pressure tactics and people who specialize in making those tactics feel good, generally Democrats but with the collaboration of all too many Republicans.

The Government Power Brokers

The developers of the ride-sharing technology will be the ones with very deep pockets. Those developers will have eyes upon federal dollars – Elon Musk’s Tesla, Google, et al. – disregarding the federal control that always comes with it. There may be less of that control with the current administration, but that remains to be seen, and any administration is temporary – the bureaucracy is eternal.

As beloved as it is of environmentalists, a tiny car like the Waymo would be intolerable on a long commute. Aside from the diminished ability to physically fit inside it, vehicle weight lends stability to the ride; a stable ride is a less physically disruptive ride and therefore more comfortable. Little, light cars are tolerable for a time, but they become increasingly uncomfortable as the ride goes on.

With a privately-owned car, at least you have the ability to break your trip and stretch your legs at the next gas station.  It seems unlikely that a self-driving Uber car will offer this ability.

This may have less of an effect on passengers who have confidence in their transport. People who have grown up with ‘transportation as a utility’ viewpoint, perhaps in major cities where most trips are by public transit anyway, may be able to forget about the innate worries of not being in control of their fate.

After a time, perhaps we will become accustomed to this new level of ego forfeiture and quit worrying about minor mechanical malfunctions. But there must be some level of awareness of the automatic. The Uber accident is a case in point: some of the automatic controls did not work, because the switches were set incorrectly.  In other words, they were turned off.

The bureaucrat’s instinct will be to remove the ability for users to turn settings off, but off is probably the proper setting – sometimes. In the current state-of-the-art systems, where a human monitors the actions of the robot, the hardest thing to do is to remember what will help and what will confuse. No one is capable of designing a foolproof system, and we have over a century of experience in trying to do so.

The Navaids based autopilot for aircraft was originally developed by Sperry Corporation in 1912, amazingly using hydraulic controls to operate elevators and the rudder. A gyroscopic heading indicator and an altitude indicator were tied to the autopilot for control over those two critical aspects of the flight.  Yes, the first autopilot was only a little more than 10 years after the Wright brothers first flight.

From its inception until the time this writer worked on Air Force Navaids was less than half a century. The Navaids of the time were at a similar state of development as the auto-driver is now.

Likewise, we have had Cruise Control for decades, but the first serious attempts at autonomous driving date back only to 1977 and the Tsukuba Mechanical Engineering Lab in Japan. It required special lane markers and other means of keeping itself straight, and was far from consumer-ready, but bore the promise of things to come.

Today, Audi has actually produced a marketable version of its A8 sedan that is autonomous up to speeds of 60 km/h (38 mph – hardly an excessive speed). Several other manufacturers are working feverishly on autonomous vehicles so that they may add their products to the flood that will soon hit the market. The Audi and others will be autonomous, but they will all require a ‘safety driver’ to be present in the cab of the car, capable of taking over operation if need be.

The three manufacturers listed at the top of this column intend no such limitation on the vehicles they are preparing to release. But we think that the bureaucratic will rule the day.

Perhaps the world of the utopian self-driving car is farther off than we imagine.

Thomas Anderson is a multi-state registered architect and an ex-Air Force electronic technician, who is a keen observer of the human condition.  Read other Scragged.com articles by Thomas Anderson or other articles on Business.
Reader Comments

Thomas- Your article seems to under estimate the value of self-driving cars by viewing only the threats coming from inside your vehicle. In fact, the biggest safety advantage of self-driving cars will be when ALL vehicles will be self-driving. Then you will be protected from the threats presented by OTHER human drivers due to various causes of distracted or dangerous driving that present the biggest danger to you in your vehicle. Additionally, the self-driving car will keep other vehicles safe from your own impaired driving ability due to intoxication, age/reflex, or plain distraction. Pedestrians will learn how to act in an environment of predictable behavior by self-driving vehicles. The end result will be many fewer fatalities and injuries both in and out of cars compared to today.

June 11, 2018 8:39 AM

In the early 90s, Merchant Marine ships had access to, although not all were equiped with, an autopilot system called the Total Navigator. Like aircraft, ships have had some form of autopilot for many many years but TN interfaced with the relatively new GPS for location, automatic radar plotting aids (radars that could track other vessels and plot their relative courses) to avoid other vessels, and electronic charts for chart data-basically google maps, but manually loaded with 5.25 floppies. This was state of the art and no one trusted it. We used it, but didn't rely on it. And that is where I see the danger of self driving vehicles. If as a pilot of your vehicle, you use the automated functions to ease the demand on your cognitive abilities and then apply that freed up ability to observing your surroundings-instead of constantly scanning the speedometer, you never take your eyes off the road, many fewer accidents would happen as the pilot would see things and be able to take over for any emergency situation when the autopilot inevitably fails. But, non professional operators being what they are will likely take a nap, read something, or chat on the phone, completely oblivious to their surroundings until they smash into something.

June 12, 2018 10:39 PM

@Scott Miller You are totally right in saying "operators being what they are will likely take a nap, read something, or chat on the phone, completely oblivious to their surroundings until they smash into something" but how is that different from what we have now? Lots of accidents happen that way.

The question is, will the autopilot smash up due to programming deficits more often than human drivers smash up due to attention deficits?

That remains to be seen....

June 20, 2018 5:49 PM

All indications were that the hapless pedestrian would likely have been struck by a human controlled vehicle. Autonomy is coming regardless and we will all be safer as a result.

June 21, 2018 10:30 AM

Yes, no autonomous safety idea is flawless, in fact many are downright dangerous. Recall the original airbag concept - a great supplementary system for head-on collisions, nothing more. Joan Claybrook, at the time head of NHTSA, declared that no law could be made to make people use seatbelts, so she provoked legislation for requirements for intallation of airbags as passive restraints. The immediate downside, of course, is that people became complacent that their 3 point harness seatbelts were no longer needed. The consequence was rapid, as rollover, side-impact, secondary impact, and rear collision accident injuries were not prevented by air bags. Add to that, the engineering designers did not factor small scaled people into the mix and many of them died through bag deployment decapitation. So, beware of the bureaucrat who makes policy from a lick and a promise.

July 8, 2018 12:16 PM

Another sociological consequence of policy driven legislation that might eventually happen from this scheme is unbearable traffic density. Currently, I80-94 corridor between Illinois and Michigan is regularly traveled by people on average, driving 80 miles per hour in posted 55. So, if bureaucrats legislate compliance with the posted speed limit of 55mph (wildly popular during the 70s, I might add) then the 10,000 care density over a 20 mile length of road will blossom to 14,545 cars on the stretch of road. If you want an instant simulation of what I am trying to describe, drive 55 in the left passing lane some day. It's illustrative and provocative of lethal road rage. Traffic throughput will seriously suffer and I will get a jet-pack to fly to Michigan on weekends.

July 8, 2018 12:30 PM

Dan- Your first comment is an apt caution to us all on the policy choices made by politicians. Your second comment, however, does not take into account the superior capabilities of driverless cars once they can communicate with each other. vehicles will be able to drive faster and closer together because they will all be adjusting speed in unison. Much higher lane densities will be possible with controlled merging and demerging. Not only that, all those folks in jet-packs will not be on the road!

July 8, 2018 1:15 PM
Add Your Comment...
4000 characters remaining
Loading question...