Self-Driving, But Slowly

Everybody thinks they want self-driving cars.

Waymo, developers of self driving cars, is about to launch into a new phase of development. Their test grounds are located in California, in Santa Clara County; the rest of us will be safe, for a while.

Waymo is a unit of Google, owned by Alphabet which is now the central corporation that maintains the thousands of fingers in the hundreds of thousands of pies that make up the Internet conglomerate. Alphabet does not think of itself as an American corporation, but it is certainly willing to accept the cash (Bitcoin?) that its far-flung services will demand here in the USA.

California roads have been used as test beds by Waymo for more than a decade including freeways, highways, and city streets. The venues have been Palo Alto, Mountain View, Los Altos, Los Altos Hills, and Sunnyvale in Santa Clara County so far. Heretofore, however, there has always been a human being in the car to monitor the performance of the vehicle as it makes its way around.

The new tests will be in a fleet of cars which will no longer require the services of an on-board pair of human eyes; they will be truly "driverless."

The state of California has granted approval for all this in light of the fact that Waymo has logged more than 10 million miles on public roads. Of course, the California DMV has a set of requirements - from Forbes magazine:

  • Providing evidence of insurance or a bond equal to $5 million.

  • Verifying vehicles are capable of operating without a driver and meet federal Motor Vehicle Safety Standards and are SAE Level 4 or 5 vehicles.

  • Confirming vehicles have been tested under controlled conditions that simulate the planned area of operation.

  • Notifying local governments of planned testing in the area.

  • Developing a Law Enforcement Interaction Plan that provides information to law enforcement and other first responders on how to interact with test vehicles.

  • Continuously monitoring the status of test vehicles and providing two-way communication with any passengers.

  • Training remote operators on the technology being tested.

“California has been working toward this milestone for several years, and we will continue to keep the public’s safety in mind as this technology evolves,” said [California] DMV Director Jean Shiomoto.  The march of technology proceeds apace - most of the time.

We've all heard of the bumps along the road.  In March, Elaine Herzberg, a pedestrian whose autopsy showed alcohol and drugs present in her system at the time of her death, was killed by a self-driving Uber vehicle in Tempe, Arizona. Her behavior at the time of the  incident was aberrant according to the reports of the scene. That Uber vehicle was running under 'automatic' control, but there was a safety driver present: a transgender convicted violent felon who was watching a movie on "her" smartphone at the time, for all the good that did.

As is usual in the US regarding radical change, there has been a firestorm of debate about the suitability of automatic vehicles, with numerous negative coverage articles from the Wall Street Journal, Motortrend, and others.  In spite of this, the state of California DMV has issued the permit, and Waymo intends to proceed with its testing and validation of the self-driving technology.

Consumer Reports lists their concerns, but most of the press and a large number of potential  customers seem to be on the side of getting on with the march of technology. Waymo is opening up a car service in Phoenix, Arizona, another test location, with a list of 400 eager experimenters. These will be riders in the self-driving Fiat-Chrysler minivans specially outfitted for self-driving and will operate within a 100 mi.² area  at first; the area will be expanded step-by-step.

Here is one list of 13 driverless car companies. And there is another list of 10 companies which is completely different in order and in content.  This whole field is so new that even the journalists that specialize in documenting the occurrences within it have different criteria for success.

And, just for completeness sake, here is a website that describes the levels of NHTSA automation on a scale of 0 – 5. That website also describes the differences between the NHTSA (National Highway Transportation Safety Administration) and the SAE (Society of Automotive Engineers) schemes of vehicle automation.

The California DMV is demanding "Level four or five vehicles," but the two different standards – NHTSA and SAE – each require one or the other: level  four = driverless for SAE, level  five = driverless for NHTSA. NHTSA is a government standard and SAE is in industry standard, so we expect that the government will prevail.

This column has dealt with a few of these automation solutions in the past, and we will try not to repeat ourselves. However, a salient point was made in that article: the first autopilot for aircraft was built and put into service by Sperry Corporation in 1912, barely 10 years after the Wright brothers had their little experiment in Kitty Hawk.

The development of technology in 1912 faced far fewer constraints than it does now; the problems that are being solved with self-driving cars are vaguely similar to the problems faced in developing an autopilot system. Sperry had simpler problems that were all mechanical, and thus could be dealt with quickly. Now, there is the bureaucracy and the restraints upon the system of having to contend with other vehicles scurrying about on the same plane (road) as the subject vehicle, but the problems are similar in concept to a designer of systems.

Contention with traffic, making the vehicle conscious of the road it is on and where turns may be made, and above all SAFETY, SAFETY, SAFETY differentiate the autopilot solution needed for an early aircraft and a vehicle traveling autonomously with hundreds of other vehicles driven by humans of all levels of competence on a public road.

What will be the future? The economics of automobile use and ownership is greatly in question. Will we continue to own our cars? Will we have a car appear at our front door, get in, and begin our work day as the vehicle drives us to an office or to some other venue where our job is required? Will we operate with some combination of the above and a commuter train? These are imponderables at this point.

The problem that the developers of autonomous vehicles face is mind-blowing. The state of California is treading where angels fear to tread. Of course, no one has ever successfully sued an angel.

One remembers George Jetson leaping from his glass canopied self-driven, collapsible car-in-a-briefcase in hand as he called out to his wife, Jane that he was home. His boy Elroy…, etc.

An amazing amount of the gadgets and doodads depicted in the Jetsons have come true, and we are on the verge of the self-driving part of that scenario. the pace of change is such that the expectations of the people involved in doing all this are in the ascendancy: there are no bounds.

First there will be a few self-drivers, then will come the rush of copycats and then the wagon jumpers which will convince the curmudgeonly few to come on board. In implementation, self-drivers will become ubiquitous slowly at first, but like every other behavioral change brought about by technological innovation, the change will eventually be accepted and adopted by all.

The fact of this change is apparent. The beginning has already happened. The only question left is how long is this going to take? For the technology required, the change will happen rapidly, but it is the societal acceptance part which is in question.

Building the Future?

For a comparison, let us look at the timetable of two New York City buildings: the Empire State building and the Freedom Tower at the old World Trade Center.

There are factors in this comparison that are not fair to the Freedom Tower building, but then, maybe that is part of the point. The attack that brought down the twin towers happened on the 11th of September, 2001. There are all kinds of societal issues which concern the loss of this landmark, buy we'll pass over those in this article.

For political and other reasons, the cornerstone for the "Freedom Tower" was set in place on 4 July, 2004. The wrangling and bureaucratic difficulties delayed design of the building for nearly two years, and actual construction began sometime later. Wikipedia gives 18 November, 2006 as a start date, and 17 May, 2008 as the day that the building reached ground level.

The structure was "topped out" on 10 May, 2013 and was opened for tenants on 3 November, 2014. The general public was allowed into the building on 28 May 2015.

For our purposes, let us include all the wrangling, and give as dates of beginning construction of the Freedom Tower as 4 July, 2004, and the date that it opened to tenants, 3 November, 2014. Therefore, the building was under construction or bureaucratic wrangling for 10 1/2 years, or longer than the 9 years it took to get to the moon.

The Empire State Building Corporation was formed for the purpose of building the building in 1929. On 17 March, 1930, construction was begun - a few months after the greatest financial crash in history. On 1 May, 1931 the Empire State Building opened officially. Project total elapsed time was 2 years and a few months; construction time was 1 year 6 months.

In months: 18 versus 126  – Seven times as long.

Why does this matter in a discussion about cars?

The discussion is no longer about cars. The issue is public acceptance and public approval. That is the same evolution that has come about with the Freedom Tower.

The world has changed, making simple architectural/engineering decisions into monumental efforts of aggregate will. Henry Ford would have died a pauper under these circumstances. Well, Henry Ford would have had to augment his assembly line in some ways that took employee health and welfare into greater account, and his attitude of  "any color the customer wants as long as it is black" would have slammed the factory doors on production.

Under those circumstances, the idea that the company Waymo can prevail after only 10 years and can set a driverless vehicle in our midst is a laughable, naïve precept. Bureaucracy has yet to take its toll.

The state of California is willing, and the state of Arizona is eager, but what about Oregon and Washington? What about the cash strapped Illinois or New Jersey? Those are states where bureaucrats determine the public good, and there isn't much of that. Other states will get into the regulatory act with a crazy quilt of requirements, also.

And if the Ninth Circuit Court gets involved, no telling what may happen, or in what lifetime.

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 articles by Thomas Anderson or other articles on Business.
Reader Comments

No, thank you!

August 18, 2019 3:03 PM

But politicians can be bought, and Alphabet has the money, and the political power to tilt elections. Forget the "Mandate of Heaven". Google intends to run the contry.

August 18, 2019 4:28 PM

"The issue is public acceptance and public approval."

Programming for safe driving is more of an issue. Social media and mainstream news can take care of public opinion in a few years.

"In implementation, self-drivers will become ubiquitous slowly at first, but [it] will eventually be [...] adopted by all."

I agree, and therein lies the danger.

Google spins the truth about current events. Try using them to find news or video of a hot-button event in progress that *isn't* part of the narrative.It's difficult. They make conservative people and websites disappear from search results. They target people who don't mimic the narrative and with the click of a mouse, they shadowban and/or wipe a content provider's ability to be seen and earn a living... and there's no recourse.

And this is the company whose software will control self-driving cars?

Google's moving beyond controlling what we see and find on the web, and adding eventual control over where we can physically go in our day to day lives... and when, and if, we get there.

August 21, 2019 2:00 PM

As an engineer myself, I find the concept of driverless cars exciting, BUT ...

Were the cars to drive on roads with neither pedestrians nor human-controlled vehicles allowed, I would feel a lot better about the safety of the system. It is one thing to design and construct a distributed-control system in which the vehicle controllers, road sensors, and communications processors all interact with one another to maximize safety, and only then efficient travel, but human traffic on the same roads, without *ANY* possible communication with those vehicles by the driverless ones, is asking for a disaster.

If enough people switch over to the driverless cars before a disaster occurs, the human-controlled vehicles will become *much* less of a problem. Presumably, the controls will improve over time, so once such vehicles really are ubiquitous, I will then - and *only* then - switch to one from a manually-controlled automobile. I'll let other people debug the system with *their* lives.

August 21, 2019 8:59 PM

Just think, the Epstein of the future could just have a driving accident with no sign of electronic failure. He just decided to drive off the cliff. The electronics clearly indicate he was in manual control when he committed suicide. That the autopsy that indicated otherwise can probably be explained by human error. He clearly wanted to die. His actions to smear our beloved President of "We know everything" Inc. proves he wanted to die.

August 23, 2019 10:01 PM

Your analogy, buildings compared to driverless cars, seems odd to me. Would it not be better to look at the acceptance of electric vehicles relative to internal combustion powered vehicles?

Back in 1996 my company offered to "give" me a GM EV1. I turned them down. Later, Georgia tax law would have made owning an electric Leaf practically free. I also turned that down since I was routinely making trips that far exceeded that vehicles range. It's now 2019 arguably 25 or more years into the process. Electric vehicles account for about 2% of vehicles sold in the US, and that's with the substantial and involuntary assistance of your neighbors who contribute to your purchase via force of government.

Obviously there are differences. If an electric vehicle takes days to make a drive that an ICE vehicle can make in a fraction of that time, that's an unavoidable problem. If your self driving vehicle can't drive in the rain, you can take control and still get to grandma's house for Christmas dinner... So, self driving vehicles should see quicker acceptance.

But you do bring up a point. Bureaucracy will be a problem and add to that the legal issues that are bound to arise once we see a few accidents. What happens when a self driving car full of kids going to school drives into a crowd of first graders getting off their buss?

Personally, I'm hoping that when i'm too old to drive I will be able to get to the doctor's office in my own vehicle by just telling it where I want to go. Until then, I really enjoy the thrill of driving my Porsche, and that includes the great sound of that engine working to accelerate through the gears as I shift my manual transmission.

September 5, 2019 8:35 AM
Add Your Comment...
4000 characters remaining
Loading question...