Need help? Text 281-761-6097
The subject of autonomous cars can be pretty confusing.
When most people think of autonomous cars, they picture a vehicle that can drive them around in their sleep with absolutely zero human intervention. Obviously, that type of vehicle doesn’t yet exist.
On the other hand, guys like Elon Musk often talk about autonomous cars as if they already exist and will be available to the public very soon.
So which is it? Do we have autonomous cars or not?
We need to dig a bit deeper. As we dig, two things should become abundantly clear:
It’s important to understand that, when speaking of cars, there is a spectrum of autonomy. Industry experts have identified six different levels of autonomy, and while we haven’t reached the level of total autonomy yet, we have made tremendous progress.
The levels of autonomy are as follows:
Most cars, especially older ones, don’t have any automation. They offer cruise control to help drivers maintain speed, but that’s not autonomy. Cruise control is set by you and doesn’t change unless you take action.
Most newer cars offer very low level automation in the form of driver assistance. Adaptive cruise control uses radar or cameras to monitor cars in front and behind, always ensuring a safe distance between your car and those around you. When traffic slows, the car also automatically slows, and it resumes speed once traffic picks up.
If you start to veer out of your lane, lane keep assist will move you back into it, which is especially helpful if you’re drowsy. However, both adaptive cruise control and lane keep assist still require the driver to be in full control of the vehicle.
Cars with Level 2 automation can handle steering, braking, and acceleration at a higher level than Level 1 vehicles. Though the driver must still pay attention and be ready to take over at any minute, there are certain situations in which Level 2 vehicles can control the basic elements of driving.
For example, the Tesla Autopilot system allows automatic steering as long as there is a divider separating cars moving one direction from oncoming traffic, but only up to specific speeds.
Other examples of Level 2 automation include General Motors Super Cruise, Mercedes-Benz Distronic Plus, and Nissan ProPilot Assist.
These vehicles are yet one step further down the road when it comes to automation. In specific situations, like on highways, they can observe their surroundings, move between lanes, and control steering, braking, and acceleration.
The first production car to come equipped with Level 3 automation is the Audi A8. In slow traffic (up to 60 km/hour), the AI Traffic Jam Pilot can control steering, starting, braking, and acceleration. When the system is maxed out, the driver is prompted to take over. The car is not yet available in the United States due to the complicated legal system and implications.
Vehicles with Level 4 automation truly are self-driving. The car can handle starting, steering, braking, acceleration, and even parking, all in a large variety of settings. In normal conditions, the driver can put the car into autonomous mode and then take their eyes off the road.
If the vehicle encounters something it doesn’t know how to process, it will prompt the driver to take control. However, and this is critical, even if the driver doesn’t take action, the car will still continue to drive and maneuver automatically.
The Waymo self-driving car operates at this level of automation. Currently, Waymo operates self-driving taxis in both Phoenix and San Francisco, and the results have been relatively positive so far. Feedback from 10,500 rides shows that approximately 70% of riders rated their experience as five stars. Most of the time, a human safety driver sits behind the wheel, ready to take over should anything go wrong. However, there have been some instances in which totally driverless cars have been used.
Honda has announced that they hope to have a Level 4 vehicle by 2026. Lyft and Uber have been moving toward Level 4 for some time now.
There are currently no production cars with Level 4 automation.
Vehicles with Level 5 automation need no human help. They don’t need a steering wheel, pedals, or even a driver. They are fully automated and can handle all elements of driving under any conditions, regardless of whether a human is present.
Nuro has been working together with Krogers on very small vehicles with Level 5 automation. The vehicle is loaded with groceries, delivers the groceries to your house, and then drives back to the store, all without any human intervention.
So when will we have fully autonomous cars? The answer is that nobody is quite sure.
Elon Musk declared that fully autonomous technology would be completed by the end of 2019, but this doesn’t mean that autonomous vehicles will be available to the public at any time in the near future.
Currently, Tesla does offer what they call “Full Self-Driving Capability” with their cars. This means that the vehicles have the capacity to automatically:
However, on their website they note:
The currently enabled features require active driver supervision and do not make the vehicle autonomous. The activation and use of these features are dependent on achieving reliability far in excess of human drivers as demonstrated by billions of miles of experience, as well as regulatory approval, which may take longer in some jurisdictions. As these self-driving features evolve, your car will be continuously upgraded through over-the-air software updates.
In other words, Teslas can perform these actions but they are not currently enabled to.
There was talk that GM would deploy self-driving electric cars in 2018 in conjunction with Lyft, but that never materialized. It was then reported that GM was pulling back and didn’t intend to launch self-driving cars until some time after 2019. According to GM, more testing was needed.
In 2017, Ford Motor CEO Mark Fields declared that Ford would have a Level 4 vehicle by 2021, with no gas pedal or steering wheel. Since then, Ford has backed off this prediction, saying that it still plans to release a self-driving car but that its applications will be very “narrow”.
So why is it so hard to predict when fully autonomous cars will be available to the public? Because there are numerous complex factors involved.
First, there is the technology itself. In order for a vehicle to be fully autonomous, it must be able to drive in a huge variety of conditions. In sunshine, rain, sleet, and snow. In congested cities and empty, rural areas. On open highways, small back roads, and in the midst of traffic jams. It has to be able to process all manner of external objects, including pedestrians, bicycles, animals, and clutter on the road. Simply creating a vehicle that can process all these different variables is an enormous endeavor that involves inventing algorithms that don't currently exist.
Then there is the issue of testing the technology to ensure that it works properly and is safe. It’s one thing to test cars in closely controlled environments, it’s something else altogether to test them in the real world. What makers of driverless vehicles are discovering is that driving is more complicated than they realized. Argo AI CEO Bryan Salesky said:
You see all kinds of crazy things on the road, and it turns out they’re not all that infrequent, but you have to be able to handle all of them. With radar and high-resolution cameras and all the computing power we have, we can detect and identify the objects on a street. The hard part is anticipating what they’re going to do next.
Finally, there are many sticky legal issues that must be resolved before driverless cars can be made available on a widespread basis. For example, what if an autonomous vehicle hits a pedestrian? Is the company responsible or the owner of the vehicle? Should the AI in vehicles prioritize the safety of those in the vehicle or those outside? What happens when two autonomous vehicles collide? Numerous states have already passed legislation that governs self-driving vehicles, as well as the U.S. Department of Transportation and NHTSA.
These regulations, while absolutely necessary, significantly slow the rate of progress at which autonomous vehicles can be developed.
Bottom line? We’ve got a long way to go before truly autonomous Level 5 vehicles are available on a widespread basis.
According to a survey by J.D. Power, most auto and tech industry experts think that it will be at least twelve years before fully autonomous vehicles are available to individual buyers.
The driverless technology that currently exists is really quite astounding. While it’s certainly not flawless, we’re certainly approaching the widespread advent of autonomous cars.
There’s no doubt that autonomous vehicles are the future. Billions of dollars are being spent every year to ensure that autonomous vehicles become a reality.
The problem, however, is reality itself. It’s incredibly, staggeringly complicated. Despite all the advances that have been made in driverless technology in recent years, we’re still some ways off from having a vehicle that can automatically handle all that’s involved in driving.
When thinking about the future of driving, tempered optimism seems to be the right perspective.