Say's Spotlight Say Insights Say's Spotlight Current Page

Why Self-Driving Cars Won’t Be Here as Soon as You Thought

Why self driving cars wont be here as soon as you thought header

By Andy Jensen on January 14, 2021 in Say's Spotlight

Pick any random sci-fi movie predicting the future, and you’ll either see flying cars or self-driving cars. We’re still decades from a flying car, but a 2015 article in The Guardian predicted self-driving cars would be on the road by 2020. So, where are they? While 2020 ruined many things, it’s not to blame here. Let’s have a look at the status and future of the self-driving car.

Why Self-Driving Cars?

Outside sci-fi movies, self-driving cars already work in the real world with limitations. Ideally, when a car is full self-driving, or fully autonomous, it can safely and efficiently take you to your destination without human interference. You can get in, text, play a game, read, or even sleep, and the car and surrounding traffic won’t care or notice as you commute and arrive at your destination without incident.

Self-driving cars sound awesome, but companies have spent more than $80 billion on the technology for more than its cool factor. Several positive aspects make autonomous vehicles attractive to a wide range of buyers. For instance, they could reduce vehicle accidents, increase driver productivity, reduce traffic congestion, and enhance mobility for older people and people with disabilities.

How Do Self-Driving Cars Work?

Have you heard artificial intelligence (AI) is the next big thing? AI dominates discussions in everything from social media to agriculture and even self-driving cars. Autonomous cars need AI to be intelligent and adaptable to the various driving situations they encounter. The autonomous car overcomes dynamic driving challenges through a complex network of computers and sensors, including radar, lidar, GPS, and visual cameras that mimic vision. The vehicle must interpret what the sensors “see” and take the correct action. It’s more than navigation as AI needs to handle traffic, weather conditions, distracted drivers, unpredictable bicyclists, jaywalking pedestrians, thrill-seeking squirrels, ubiquitous road construction, tire-destroying potholes, and the random cardboard box in the road. That’s a lot to ask of software, and manufacturers’ various autonomous offerings can get confusing quickly.

Fortunately, the Society of Automotive Engineers (SAE) stepped in to help bring clarity to automation by establishing levels for autonomous vehicles based on how a driver interacts with the vehicle.

  • Level 0

These vehicles have no autonomous features. This level covers everything from the first 1885 Benz Patent-Motorwagen to every vehicle built through 1998, and most vehicles even today. Cruise control doesn’t count, as it’s not an adaptive system; it operates at the driver’s setting, much like the headlights or radio.

  • Level 1

These vehicles have “driver assistance autonomous features.” Mercedes-Benz pioneered another first 115 years after the Patent-Motorwagen when their 1999 S-Class included Distronic, a retro-sounding name for the world’s first assisted driving tech. Radar measures the distance to the vehicle ahead, and the computer adjusts the speed to maintain a safe distance. Level 1 automation means the driver is in full control of the vehicle but assisted by the tech.

  • Level 2

Most “self-driving” cars fall into this category today, which is technically “partial automation,” according to the SAE. The driver can let go of the steering wheel for several miles but must pay attention to traffic conditions. Tesla’s Autopilot, BMW’s Personal Copilot, and Cadillac’s Super Cruise have level 2 automation.

  • Level 3

The next goal seems hard to hit. Publicly available vehicles have been stuck around level 2.5 for the last few years. Level 3, also called “conditional automation,” allows more self-driving than level 2. The vehicle understands to slow down or move for a disabled vehicle, or to pull over for emergency traffic. This means you could turn your attention away from the road and have a face-to-face conversation with your passenger. The tech seems to be capable, but so far manufacturers have promised and then shelved Level 3 cars.

  • Level 4

Full-self driving happens at this level, but the technical term for it is “high driving automation.” The vehicle can tackle nearly any situation, but the human driver can take over when desired. Picture Will Smith in I, Robot letting the car drive at first, but driving when he feels the need. These cars are already on the road in a limited way. The company Waymo operates a Level 4 taxi service with no drivers in the Phoenix, Arizona area.

  • Level 5

With full driving automation, drivers may not have the option to drive. Level 5 vehicles may lack steering wheels, pedals, or even driver information (vehicle gauges) because they don’t need them. You get in, go to sleep, and the car takes you to work. This is the likely future because it delivers the biggest benefits. For instance, it allows older people and people with disabilities to reclaim independence and frees parents from being kiddos’ taxi driver.

Which Self-Driving Cars Are Available Today?

Want to buy a self-driving car? You have limited options for purchase, but some high-tech alternatives are on the road.

  • Audi, BMW, Cadillac, Tesla, and other mainstream manufacturers sell vehicles with Level 2 semi-autonomous technology. The technology adds a few grand to the price of the vehicles and isn’t available on every vehicle or trim level. It’s also not true self-driving, but future software upgrades could promote these vehicles to Level 3 or 4.
  • Waymo is the only company transporting passengers in Level 4 self-driving vehicles without a human co-driver. However, the vehicles are “geo-fenced,” meaning they only operate in a small area mapped by the vehicle’s software. Want to leave the company’s 100-square-mile geo-fence surrounding Phoenix, Arizona? Like Hal 9000 on 2001: A Space Odyssey, your robo-taxi is likely to say “I’m sorry, Dave. I’m afraid I can’t do that.”
  • The robotics company Nuro, founded in 2016, is already ahead of the crowd, receiving DOT and NHTSA approval for their autonomous R1 and R2 cars. Nuro’s key to speed? They won’t carry any passengers. Instead, their small delivery cars can deliver groceries or food delivery orders. The R2 is continuing testing for select customers in Houston.
  • Taking the slow lane could pay off as robotaxi start-up Voyage aims for a top speed of 25 miles per hour while transporting older people around two retirement communities in Florida. All Voyage taxis have human backup drivers for now, but the company hopes to eventually launch Level 4 service nationwide, even in the fast lane.
  • Embark is moving heavy freight in robo-semi trucks and has already completed an impressive coast-to-coast trip. Everyone from Amazon to Walmart is pulling for autonomous long-haul trucking, with companies such as Nuro taking freight the last mile. The progress is slow though, and long-haul truck drivers are secure in their jobs for now.
  • Daimler and Tesla have quietly dropped their plans to put thousands of self-driving taxis on the road by 2021. The massive technology development expenses coupled with slower than expected technology breakthroughs mean companies must push their plans another few years, not a few quarters, down the road.

The Feds are unimpressed with the current autonomous vehicles on the road. The National Transportation Safety Board Chairman, Robert Sumwalt, said in a press release in February 2020, “There is not a vehicle currently available to U.S. consumers that is self-driving. Period. Every vehicle sold to U.S. consumers still requires the driver to be actively engaged in the driving task, even when advanced driver assistance systems are activated.” It turns out building self-driving cars is hard.

What’s the Holdup with Self-Driving Cars?

Machine learning, a form of AI, is the key to successful self-driving vehicles. AI can have all the sensors and processing speed in the world, but it’s useless without the wisdom to back it up in unforeseen circumstances. AI needs a lot of practice, covering millions of miles, to be equivalent to a 16-year-old who passed a driver’s test last week. Self-driving vehicles have covered billions of miles in test simulations and millions of driverless miles on the road. Many companies drove those miles, but don’t typically share their experience with each other. This is why Waymo’s taxis still need human assistance an average of once every 13,000 miles. Drivers want self-driving cars, but we need them to be safe.

You may have seen clickbait headlines about how hackers can take control of your self-driving vehicle. All connected vehicles, not just autonomous ones, are potentially at risk. Your Bluetooth, Apple CarPlay, OnStar, and 4G Wi-Fi hotspot are all connectivity points to your ride. Hackers could break into current vehicles and mess with settings or find financial info if you’ve used it in your vehicle. Autonomous vehicles step up the potential mayhem because hackers could take control of vehicle systems, for instance suddenly activating the brakes or hijacking it to drive to a different location. While it sounds ominous, many cybersecurity experts say car hacking likely won’t be a big risk due to the multiple layers of sensors and security they’d need to overcome.

Finally, while it’s tempting to blame 2020 for not living up to expectations, COVID-19 also deserves some blame for the lack of autonomous cars. The worldwide pandemic shut down or slowed engineering on autonomous vehicles. But at the same time, it elevated the need for them. Public transportation has taken a hit because it exposes people to the virus, and self-driving vehicles provide a safe mobility solution for those who can’t drive themselves.

Hold On To Your Steering Wheels, For Now

Because the potential payoff could be large, companies and investors are pouring billions of dollars into the self-driving car industry, creating the tech infrastructure for a new transportation category. Nearly everyone will be a customer, because removing the human factor from driving theoretically makes for safer vehicles. This means saved lives, saved time, and less legal wrangling. Plus, individuals and fleet companies may save money. Everyone may win in the robo-car future. But despite what you may have heard, it’s not happening next year.

Andy Jensen is a consultant for Say Insurance. He's an automotive enthusiast writer specializing in new and used models, industry tech and trends, and the car culture that surrounds it all. After receiving a Bachelor of Journalism from the University of Central Oklahoma, he decided to write about cars instead of getting a real job. He’s written for Jaguar, Volvo, Ford, Advance Auto Parts, Haynes Manuals, and others. His project car probably isn’t running.