Travel & Adventure

Sorry, but Autonomous Cars Won't Be a Reality Any Time Soon

There it is. The question is forming in their mind. They’ve learned I write about cars and it’s coming. Young, old, male, female; no matter the variables, people always want to know my take. “Autonomous cars are right around the corner, right?” They ask. “Maybe five years?”

In that moment, I nod and listen to their anecdotal and often misinformed evidence. They talk feverishly about Tesla’s Autopilot. Maybe Uber or Waymo if they’ve read anything past the dissertation’s worth of wordage delivered by numerous authors across the media on Musk’s cult of personality. And everyone is convinced our “his dog Elroy” future is mere months away.

They then stop talking and finally look directly at me and ask, “So what do you think?” I reply with a simple and succinct, “Nope.” Their mouths agape, brows furrowed. They huff and often leave before I can explain. But as much as I’d like to reaffirm their beliefs—self-driving cars have the real possibility of reducing the number of automotive fatalities—it’s just not going to be in our near tomorrow.

Our autonomous future has been just around the corner since the 1960’s space age when our Jetsons-like life was all but assured. Automotive concepts, advertisements and marketing campaigns heavily used the dawning technological revolution to push futuristic technology that captured the public’s consciousness and had them buy into this narrative. After all, how hard could it be to build an autonomous vehicle in the age that saw humans traveling to the moon, beginning the nuclear era and blistering our eardrums to Hendrix’s transcendental guitar-shredding skills. Still, those spacey concepts never materialized and it isn’t hard to see why given our automotive technology was still primitive—seat belts were still optional for Christ’s sake.

While it’s not the sort of "fake news" our President likes to disseminate, the media has often delivered misguided information to the public on what autonomous and semi-autonomous systems actually represent.

It's only recently that the prospect of autonomous automobiles has edged closer to bridging the gap between science fiction and reality. This is in large part thanks to the digital era bringing down cost and a reduction in the size of technologies such as LiDAR, radar, computers and the nascent stages of artificial intelligence that are required to make autonomy possible.

True autonomy is still decades away from the average consumer, the chief reason being that the technology just isn’t ready for primetime. While it’s not the sort of "fake news" our President likes to disseminate, the media has often delivered misguided information to the public on what autonomous and semi-autonomous systems actually represent. Case in point, Tesla’s Autopilot: an unfortunate name to describe a semi-autonomous system. Because of the aeronautical connotations behind Tesla’s naming scheme, and the often ignorant hyperbole delivered by technologically illiterate writers, there’s a good cross-section of the public that believes Tesla’s Autopilot is ready to drive you from point A to point B without driver intervention. It isn’t.

What Tesla’s current Autopilot actually represents is little more than an advanced cruise control. Tesla’s cars can—through a driver’s direct prompts to the car and its radar, ultrasonic sensors and battery of cameras—keep a constant speed, maintain proximity to the car ahead on the highway through acceleration and braking, and manage the vehicle’s position between a highway’s lane divisions. The driver, however, still needs to be in control and attentive behind the wheel, and when such protocol hasn't been followed in the past, it resulted in many Autopilot-related accidents.

In tech-speak, Tesla’s Autopilot represents only Level 2 autonomy on a five-position scale. And though Level 3 and 4 automate the process even further, these categories still require humans to remain vigilant and ready to take over if the autonomous system’s bank of sensors and computer brains fail. Fully autonomous, or Level 5 autonomy, is where the car, or mode of transportation, essentially turns the vehicle into public transit, and passengers are only required to hop on in, queue up a movie, start of the day’s work, or decide to get intimate with a fellow passenger, and select a destination.

None have managed to truly bridge the gap between a semi-autonomous and a true driverless experience.

Other companies have made similar headway as Tesla, as well as similarly aggressive proclamations about their autonomous technology—Audi recently announced through marketing materials it would make available Level 4 autonomy, but the system‘s specifications were closer to Level 3. None have managed to truly bridge the gap between a semi-autonomous and a true driverless experience. Waymo and Uber remain the biggest names, apart from Tesla, that go the furthest in terms of research and development as each company’s technology is ahead of many traditional automotive manufacturers like Cadillac and Volvo. Waymo recently announced that the company’s vehicles had crossed 10 million fully automated miles with its fleet of LiDAR and radar-augmented Chrysler Pacifica minivans. Part of that distance was completed on an abandoned Air Force base rebuilt to resemble a city, with the rest on actual public streets in Arizona.

But each company has had its share of glitches, issues, controversy, and technological barriers. Waymo, which is quite candid in its development told CNET that those 10 million miles were “about safety.” And even if that’s still the case, the company is still figuring out its next step and just what the public expects from autonomy. For example, Uber and Tesla's programs have been involved in fatal incidents.

But technology issues aside, other factors play into autonomy not yet being ready, including the dissemination of such autonomous systems to traditional manufacturers like Ford, Chevrolet, BMW, Mercedes-Benz, Hyundai and Honda. With much of the field being led by private corporations like Waymo, Uber, comma.ai and Tesla, little has been discussed on how these systems and companies will integrate with the traditional auto sector. Some companies have made internal progress with their own autonomous systems—Cadillac. And some have partnered with the aforementioned external companies—Waymo and Jaguar Land Rover.

But the selling of these systems and the integration with the manufacturer’s in-house developed sensors will need sorting. The real issue will be making the proprietary software work with the manufacturer’s native sensor arrays. Essentially, if Tesla, once Autopilot is capable of Level 5 autonomy, decides it wants to sell the technology, how will outside manufacturers like Ford integrate Autopilot?

Yet the biggest hurdle, at least in my opinion, will likely become the adoption of such technology. Not just by long-established manufacturers, but by you and me, the people who actually purchase automobiles. Most technology reporters have used improper methodology when they consider the adoption rates of autonomy. More often than not, they see autonomous vehicle adoption in a similar way as iPhone adoption. Put in that context, there were early adopters—they see these as Tesla folk—then a very early tipping point that made the smartphones almost necessity to every person. So while I agree there will be a tipping point where autonomy will become a necessity, the timeframe is completely wrong. Smartphones are relatively inexpensive—they can be bought, used, destroyed and replaced with relative ease and without real financial commitment. Cars? Not so much. The median lifespan of an automobile in the United States, according to the data-gathering giant IHS Markit, is currently 11.6 years (And it’s tracking to be even higher in the next few years as wages stagnate and the cost of living increases.). Automobiles, unlike phones, are long-term depreciating assets.

Put plainly, even if the technology could be ready in the near future, and it was applied to every single new car available—something that most definitely won’t happen considering things like navigation, Bluetooth, pre-collision warnings and Level 2 autonomy are still options—people will still be driving their older cars for decades to come before they decide to upgrade to a Level 5-capable car.

The soothsayers, augurs, futurists, and much of the media that has bought into the autonomous revolution being just around the corner, are, like the supposed clairvoyants of Greek mythology; wrong. Sure, real Level 5 autonomy has the potential to come in the near future, but your ability to purchase a capable machine—outside your average production car—is still likely two decades away.

That’s not to say we should be no less fascinated by our autonomous future nor any less scrupulous in our evaluation of these technologies. Especially since we, as humans, tend to be easily misguided by language used by companies, the media, and your friend Brenda down the street. The fact of the matter remains, there isn’t a single autonomous car on sale today—and decades later, Mr. Jetson's car is far from a reality.

Related Topics