Self-driving vehicles, also known as autonomous or driverless vehicles, are the automotive industry’s biggest up and coming technology. These vehicles use computers and sensors to completely replace human drivers on the road. Imagine being able to drive to work - without ever having to actually drive.

While most think of these autonomous cars as being a distant technology better suited for a sci-fi movie, humanity might be closer to this reality that you’d expect. In fact, it’s just about time to stop wondering when technology will be advanced enough for these cars, and to start wondering how quickly society can adopt these machines into day-to-day life.

Did you know that Waymo, a subsidiary of Alphabet (Google), has already driven over 4 million miles on public roads with its fleet of self-driving vehicles? This doesn’t even include the billions of miles that have been simulated by supercomputers in order to improve performance and safety!

How do self-driving vehicles work?

When you drive your car, you are performing many tasks all at once to get from point A to point B safely. Whether or not you realize it, you are taking in hundreds of stimuli, performing your own light-speed calculations, and making real-time decisions. Simultaneously, you could be:

Observing your speed and comparing it to the speed-limit sign you just saw
Acknowledging that the car behind you is following a little close for your comfort
Noting the distance between you and the truck in front of you
Wondering if your engine is still making that weird sound from last week
Preparing for the right turn you need to make in a mile

From this, you make a decision that you need to move into the right lane and reduce your speed. This may seem like a simple decision, but you actually just combined that list of observations with your knowledge of the rules of the road, your familiarity with how your car handles, and your past experiences in similar situations. When you put it all together, you just made quite the complex decision.

Self-driving vehicles need to be able to make that same decision - So how do they do it? The first step is gathering data. These vehicles need equipment on them that act the same as your eyes and ears do.

Position sensors: A combination of a global positioning system (GPS), altimeters, tachometers, and gyroscopes allow the car to know precisely where it is.

Radar sensors: Embedded at the front, back, and sides of the vehicle, radar allows the car to know what objects are around and how close they are. This includes things like other cars, cyclists, and guard rails.

Video cameras: Coupled with advanced image processing, video cameras help the car read road signs and detect traffic lights. They also combine with the radar sensors to more robustly identify objects in and near the roadway.

Light Detection and Ranging (Lidar): Lidar sensors are mounted on the top of the car and bounce pulses of light (lasers!) in all directions to help the car identify lane markings and other road edges like curbs.

All of the above sensors are fed into a central computer onboard the self-driving car that processes all of the inputs and makes real-time decisions based on past experiences - Just like you did!

What is the current status of self-driving vehicles?

SAE (formerly The Society of Automotive Engineers), the leading international consortium on transportation, released guidelines for autonomous vehicles in 2014. These standards outline 6 different levels of automation for on-road vehicles.

SAE levels 0, 1, and 2 require human drivers to be in control of the vehicle. Levels 3, 4, and 5 are where the automated driving systems are in control of the vehicle, with humans involved in varying degrees of interaction.

Luxury vehicles have included some lane-assist and emergency breaking/accelerating for a few years now. One notable example of this is Tesla’s auto-pilot feature, which puts the car squarely at SAE Level 2.

In terms of commercial adoption, this is where we currently are. You can walk into car dealerships today and have numerous options for Level 2 vehicles available for purchase.

What is the immediate future of self-driving vehicles?

While you certainly don’t see them everyday, SAE level 3 and 4 vehicles already share our roadways. While I previously mentioned Waymo’s 4 million autonomous miles driven on public roads, there are actually dozens of companies exploring these higher-level self driving vehicles. In fact, most major car companies have expanded their research into the field, which is great evidence for how the auto industry as a whole is valuing the importance of autonomous vehicles.

In the spring of 2017, Navigant Research published a paper ranking these companies by who they expect to emerge as the industry leaders. Based on a set of 10 criteria encompassing both strategy and execution, Ford and GM come in at the top of their list. However, hidden among the list of mostly established automakers include some promising younger companies such as Waymo, Delphi, Uber, and Tesla.

All in all, the abundance of competition means we will be driving around in self-driving vehicles sooner, rather than later. Consensus among top manufacturers seems to huddle around the 2020 mark for commercially available autonomous cars. However, some companies have expressed more ambitious timelines:

Last year, Tesla began making sure that all of their cars have the necessary hardware for full autonomy, even if the software isn’t fully ready. Elon Musk even publicly announced that by the end of 2017, he will showcase a Tesla vehicle driving from Los Angeles to New York City without ever needing human interaction. He has since delayed the official demonstration of this into early 2018.
Uber has begun a pilot program where users in a small piece of downtown Pittsburgh are being driven around in self-driving cars. Currently, a human operator still sits behind the wheel in case anything goes awry.
Similarly, GM is rumored to be deploying thousands of self-driving vehicles in 2018 for ride-sharing purposes alongside Lyft.

Expert Forecast

To get some thoughts from an industry expert on self-driving vehicles, I reached out to David Silver, who leads the curriculum for a Self-Driving Car Engineer course at Udacity. David was formerly part of the autonomous vehicle team at Ford as a research engineer.

How long do you think it will be before we see an SAE Level 5 vehicle that is commercially available to consumers?

Level 5 vehicles will take quite a bit longer than Level 4 vehicles. Level 4 vehicles are already able to operate in small geo-fenced areas, such as downtown Pittsburgh, suburban Phoenix, and Singapore. Those vehicles require high-definition mapping, which is why the geofences are currently fairly small, and in fairly dense areas. To build a Level 5 vehicle that can drive anywhere will require finding a solution to either the need for or difficulty of creating high-definition maps for self-driving cars.

How long do you think it will be before the majority of the cars on the road are fully autonomous?

The average lifetime of a car is ~11 years. In dense urban areas, self-driving cars may become a realistic transportation option within the next 5 years. If we assume self-driving cars replace human-driven cars as they wear out, then it might take 10-15 years to see self-driving cars as the majority of vehicles on the road. And that is only for dense urban areas in the United States. In rural areas, and internationally, it will take much longer.

How will our roadways look different than they do now?

In 10-15 years, I expect we will see signs and signals that are designed with self-driving cars in mind. It could be as simple as a sign that displays its precise geolocation, to help autonomous vehicles localize themselves. We maybe begin to see some vehicle-to-infrastructure communication, whereby autonomous vehicles can communicate their intentions to traffic signals. Some highways might have autonomous vehicle-only lanes that have higher speed limits. But even in the 10-15 year timeframe, I expect self-driving cars will have to successfully navigate roadways that are designed for humans.

In what significant ways do you think our everyday lives will be improved because of self-driving vehicles? What might a day in the life look like for someone who has adopted autonomous car technology?

I think self-driving cars will change our lives in ways we can't even imagine. 1.3 million people are killed in auto accidents worldwide each year. The WHO and NHTSA estimate that between 3% and 6% of GDP is lost every year due to automotive accidents. Think of how much economic growth we will get back if self-driving cars can save even half of those lives!

I foresee autonomous vehicles having a huge impact in real estate, food service, shopping, childcare, hospitality, and logistics. Imagine getting in a self-driving car at your house in the countryside. The car whisks you to your job in the city, while you work. In the evening dinner gets delivered by a self-driving delivery service, while you hop into a new self-driving car to sleep while you travel on a business trip to a new city.

Besides self-driving vehicles, what bleeding-edge technology are you most excited about?

Flying Cars!