Driverless Cars

Don’t hold your breath.

It was the summer of 1970 and it was memorable for many reasons. Numerous beloved entities were cruising toward the end of their lines, while an unknown was preparing the invention that would change our world. Before the year was out Jim Morrison would quit The Doors, while The Supremes, Simon and Garfunkel and the Beatles all would officially split up. While American Top Forty was experiencing turmoil, Federico Faggin would join Intel; and just as 1971 officially began he had created the first commercially available microprocessor, the Intel 4004.

From there all things computerized would come, including today’s promise of self-driving vehicles in five or 25 years depending on the source one believes. Maybe it was fitting that The Beatles’ last No. 1 song was “The Long and Winding Road” — as apt a description of the struggle to get to autonomous cars as any.

The point has been made, most recently in Gordon Dickson’s excellent column in the May 18 Star-Telegram, that so many of the things being touted today as “great new automobile safety devices or driving assists” are already options in our automobiles. My experience so far is that they just don’t work as well as they could. A case in point might be vehicles that are capable of parking themselves, an option first introduced on the Lexus LS sedan back in 2006. At the time Fox Four was excited about demonstrating that feature live on Good Day. After all, that self-parking Lexus was the brand’s magazine and TV ad campaign at the time. So I placed a call to Toyota’s PR firm to arrange for an LS equipped with that feature to review — and immediately they begged me not to do it. The reason? The system didn’t work properly.

In the end Steve Noviello got an LS from a local dealer and live, on the air, had the car parallel park itself. It didn’t make it within 18 inches of the curb, and it wasn’t parallel to it, either.

On May 11 of this year, Jared Overton of Utah summoned his Tesla Model S with the autonomous parking feature and it drove itself into the back of a parked trailer. Tesla immediately reminded everyone that this particular feature was a Beta test; in the future, we were warned, owners should be on private property and pay attention to their surroundings if using that system. Really? On private property, you can usually park easily all by yourself; it’s on unfamiliar streets that you may need autonomous help. So, for the most part, 10 years has gone by between the self-parking LS and the autonomous Tesla, and the best advice is still, “You’d best pay attention to what your car is doing and be prepared to stop it at any time.”

To be fair, another article, this one at the ARS Technica website, claimed an almost flawless drive from Houston to Austin using the Tesla’s self-driving feature. Still, insiders claim that we will still have to resolve all of the vehicles’ other, far more complicated issues. They have grave doubts about any claims that fully self-driving cars will be on the road in four years.

One should also be reminded that a major story reporting Uber’s work on self-driving cars was published last year as their first one hit the roads of Pittsburgh. Sure enough, the AP photo showed a Ford Fusion hybrid electric with tons of lumpy gear all over the roof. Further, news stories published last year still figured that the total cost of all of the gear necessary to make a fully self-driving test car was in the neighborhood of $90,000.

Three things will have to happen to make a self-driving car feasible. First, the cost of the gear needs to come down to around $5,000. Only then will automakers make the self-driving feature a $9,995 option. Second, the gear has to be shrunk down to where it’s designed into the automobile, invisible on the outside; additionally, higher-voltage systems will have to be developed to run the computerization, sensors, cameras and wireless system that communicates with other vehicles. Finally, they need to perfect the system, because right now it doesn’t work well enough that customers would never have to worry about their vehicles’ performance under any conceivable driving situations.

I’ve reviewed the new Volvo XC-90 with its autopilot feature. Moving into downtown Dallas I engaged it, locking onto the Mustang in front of me. And for a couple hundred feet it worked. That is, until the Mustang slammed on its brakes — and I did, too, since it didn’t feel like the Volvo would stop itself before we had a rear-end collision. To be fair, I did manage a 10-mile run in that Volvo using its Lane Departure system and adaptive cruise control. But, although Volvo’s driver assist options are the best on the market today, hands down, one would hardly call that a self-driving vehicle. It was more of a stunt than a statement of functionality.

A few months before that review I had tested the Mercedes S63 AMG, also with a Lane Departure system, and this one was more advanced than one might think. It first warns you that you are drifting out of your lane, then uses the braking system to put you back dead center. The only problem was its braking and centering, even though the vehicle I was driving wasn’t straying a bit from dead center in its lane.

In fact in many ways the ultra-hype on self-driving cars reminds me of the similar push 16 years ago for hydrogen fuel-celled vehicles. Bill Ford was quoted in the November 11, 2000, issue of the London Guardian newspaper as saying, “Ford is spending $1 billion to put a family car on the road by 2004 that runs on hydrogen, emitting only water from its exhaust.” Wow, Ford was going to beat every other manufacturer promising imaginary hydrogen fuel celled cars with its own non-existent car by two full years. What that story did not say, but other articles on Bill Ford from 2000 did, such as Fortune’s Motown Cool, was that the Ford himself preferred driving an all-electric Ranger pickup truck. So it’s not like Bill Ford couldn’t see true value in a different future for automobiles.

Now, in that Guardian story, Ford also talked about the future of the auto industry as he saw it. “The company is also repositioning itself as a purveyor of mobility. The day will come when the notion of car ownership becomes antiquated. If you live in a city, you don’t need to own a car.” He envisioned a future where Ford owns the vehicles and makes them available to motorists as and when they need access to transport.

Sound familiar? It should. Fortune published another Bill Ford interview a week ago, this time under the headline, “Where Ford Motor Co. Will Be in 5 Years.” Bill again talks about Ford’s partnership with Zipcar and how they have a tech lab out in California now and will invest in other tech startups. Again that Ford would be a purveyor of mobility. Then he slammed the feds, praising local governments as being the most helpful as Ford moves into being more of a mobility services provider.

That’s right, the exact same speech and predictions he gave 16 years ago.

For the record, Ford never delivered that highly touted 2004 hydrogen fueled vehicle. But to be fair, neither did General Motors — and GM suggested that hydrogen would fuel most of its fleet somewhere around seven years ago. And after 16 years, only the most basic, minimal platform for ride sharing is available, and creating it required no new technology at all. Then again, 16 years ago the Ford Motor Company made most of its profit on sport utility vehicles and pickup trucks — and today the Ford Motor Company makes most of its profit on sport utility vehicles and pickup trucks. That’s right, in 16 years the auto industry has been talking about the same future endlessly, but at best all they have done is tread water.

True, we have added new safety features to our automobiles, such as the above-mentioned lane departure warnings and automatic braking. But they are no different from other safety inventions added to cars in decades gone by, like anti-lock brakes.

Then again, if cities, states and the federal government don’t work on fixing our roads, then self-driving cars won’t work until even farther in the future. Again, having tested many vehicles with the latest safety features, I can testify that if the road lanes are not perfectly marked with painted lines, the car can’t stay in a lane. If a stop sign is knocked down, the car doesn’t know to stop. If it rains or snows, the car can’t find the road at all. Not to mention that someone’s landscaping or roadside overgrowth can obscure the road signs and throw the car for a loop.

Even after these highly touted vehicles come to market, however, there will be serious long-term problems. I’ll bet you know what they are, if you work on a computer. Today I work on a 2011 model Mac Pro, running OS 10.8.5. There have been three major operating system upgrades since I purchased this machine. But today, and for the future, I’m frozen in computer time. There’s a good reason: Upgrade your operating system and maybe your very expensive high-speed color laser printer no longer works. Or, as in the case of the latest Mac operating system, you have to replace Pro Tools for recording and repurchase around $4,000 worth of software sound effects used in that recording software. And Apple just keeps walking away from the professional market, whether in recording studios worldwide or even in Hollywood film editing.

The same hardware/software compatibility issues will arise with self-driving cars. Software will be updated and improved and new features added, until at some point more onboard hard drive space will be required. The WiFi that enables the car’s connections to others will change. The speed of the computer chips that drive these systems will grow by far the fastest; and once they do, just like with your personal computer, software engineers will take advantage of that speed.

It becomes the ultimate in planned obsolescence: The self-driving car you buy in 2025 may no longer accept new software updates in 2035, because the onward march for faster chips, product features and software upgrades will never stop. But your vehicle’s or its operating system’s ability to accept those improvements will have an expiration date. Just like any computer any of us have ever owned.

Back in 1980 Honda created the world’s first car navigation system, but it was nothing like today’s GPS units. No, Honda created a gyro-based nav system for Japan, and it wasn’t all that hard to do. The streets in that country were already built; and the gyro was so precise that, once it was set at the end of the assembly line, one could drive all over Japan and it would be accurate within 50 feet or so of the map display. Needless to say, the country’s growth made it obsolete within a few years — just like the earliest GPS satellite navigation systems often could handle only five or six states at a time. So that first navigation system was outdated seven years later, when Toyota brought out the first CD-ROM based nav system, while the first of those breakthrough “modern” nav systems from the 1980s are worthless today.

We need to accept the fact that driverless technology is not even close to being ready to put on the market — at least, not the kind you never have to worry about when you’re going somewhere. Our roads and highways are not ready to accommodate these vehicles. And all of these incredible vehicles will become obsolete within a few years of their introduction, simply because computerization changes so fast.

Other than that, it’s a great idea.

This story was originally published May 26, 2016 at 3:40 PM with the headline "Driverless Cars."