I’ve heard one phrase many times during long, monotonous highway drives: “How did people do this before cruise control?” Allowing your vehicle to regulate its speed automatically lets you relax and give your mind a break.
Take this concept, multiply it several times over, and you arrive at self-driving cars.
The idea of resting in your seat while the car does all the work is quite appealing. However, many people are reluctant to give up control. There are legitimate downsides to self-driving cars, but many fears stem from a lack of knowledge of the technology.
Hopefully, this post helps you understand this emerging sector more deeply and decide whether you want to trust yourself in this new class of vehicles.
What Are Self-Driving Cars?
Self-driving cars are exactly what they sound like. They are cars that detect the surrounding environment and make handling decisions without the driver’s help.
Functions like speed control, steering, braking, lane shifting, and even headlight activation are managed in real-time through the vehicle’s computer.
However, most self-driving cars do not operate 100 percent independently. They require commands from passengers or a control service to get started and know where to go. There are also varying degrees of automation where some features are left to the vehicle and others to the driver.
The Six Levels of Driving Automation
The Society of Automotive Engineers (SAE) outlines six levels of driving automation from level 0 to level 5.
There’s a wide disparity in features within the first three levels (0-2), but they all refer to setups requiring the driver’s attention. These levels may include some automated services, but these are tools to assist you, such as lane centering and adaptive cruise control. You are still driving.
The last three levels (3-5) are probably more in line with what you picture when thinking about self-driving cars. These vehicles are making all the road decisions on their own. You may be able to override the programs and take over, but the goal is a hands-off experience.
How Do Self-Driving Cars Work?
Self-driving cars are equipped with a series of sensors that create a real-time map of their surroundings. They’re trained to recognize and read road features such as lane positions, streetlights, road signs, nearby vehicles, and pedestrians.
However, recognizing what’s around us is only half the battle. The most crucial part is knowing how close or far away that car is in the next lane. Autonomous vehicles emit light pulses that bounce off the surroundings and measure their exact distance from them.
All this information is fed into complex algorithms managed by the electronic control unit (ECU) to determine the most suitable response.
Now that we’ve covered how these cars “see” and respond to the world, let’s go over how they know where to go. The passenger designates a destination through a map service, primarily through Google Maps.
The vehicle then performs “path planning” to determine the most expedient and safest route. Depending on the available data, an autonomous car will consider how difficult it will be to change lanes, pass vehicles, or check for roadblocks on the way.
Who Invented Self-Driving Cars?
The idea of a fully autonomous vehicle has been around for decades. It was dreamt up in early science fiction, but the first individual to create a road-ready version was Ernst Dickmanns and his team in the EUREKA PROMETHEUS project.
Throughout the 80s and 90s, Dickmanns re-engineered an S-Class Mercedes-Benz into the first autonomous vehicle. The vehicle had a public drive in 1984 but only passed through uncluttered roads.
Its first real test came in 1994 when it drove 620 miles through heavy traffic on Parisian highways. This demonstration showed a vehicle’s potential to maintain safe distances, perform lane changes, and pass other drivers utilizing cameras.
While the EUREKA PROMETHEUS project was underway in Europe, the Pentagon was doing its own research alongside Carnegie Mellon University. Their research introduced early forms of Lidar distance measuring.
In 1995, the US Department of Defense funded the Navlab 1’s journey from Pittsburgh to San Diego. The drive was nearly 3000 miles long, and the Navlab 1 controlled steering for over 98 percent of the journey.
The excitement (and government funding) surrounding autonomous vehicle development drew teams from several universities. The DARPA Grand Challenge pitted teams against each other in a race from California to Nevada.
No team reached the finish line in the first year. However, five teams completed the route only one year later. This fervor for automated vehicle research spurred many more developments and culminated in the Tesla and Waymo of today.
Are Self-Driving Cars Safe?
Losing control over a multi-ton hunk of metal is a scary thought. There’s relatable anxiety around potential malfunctions or the underlying software’s reliability. People see hectic highways and clogged intersections and doubt the vehicle’s ability to navigate safely.
Forbes reported that 93 percent of Americans felt that self-driving cars were unsafe. However, it also reported that 81 percent of Americans had never ridden in such a vehicle.
We won’t say there aren’t cases of accidents and injury from self-driving cars. Cruise, a subsidiary of General Motors, recalled nearly 1000 of its cars after one of them dragged a pedestrian for twenty feet in an attempt to avoid crashing.
This case raised commonly cited problems with automated vehicles, machine learning, and artificial intelligence in general. Driverless cars are not capable of human ethics and follow a strict code. Cruise’s vehicle did not determine that it would be better to hit another vehicle rather than drag the pedestrian.
One of the biggest red flags was Tesla’s massive vehicle recall after the NHTSA reported many autopilot-caused accidents. The event was alarming because Tesla models composed the majority of self-driving cars on American roads.
Despite recent events, there are places where autonomous vehicles are already reporting fantastic safety metrics. Waymo, famous for its self-driving taxi services, reported that its driverless cars were 6.7 times less likely to crash than human-controlled vehicles.
The most significant distinction between Waymo and Tesla is that the former’s cars operate in limited areas. Their taxi services have access to copious data regarding road conditions, traffic, and even pedestrian density at certain times of day. The city's Waymo works to even connect the streetlights to Waymo vehicles for further safety.
Pros and Cons of Self-Driving Cars
Autonomous vehicles promise to transform our roads by offering new levels of efficiency and safety. However, while the potential benefits are substantial, there are also serious concerns and drawbacks to consider.
Pros Of Self Driving Cars
One of the benefits of a self-driving system is that the sensors don’t get tired. Their eyes don’t slowly close on dark drives, stray to their phone screen, or miss a stop sign. In terms of reliability, these sensors and programs cause significantly fewer problems than what arises from human error.
These cars may also help people with mobility issues, such as the disabled or elderly, get around. These people often rely on family members or live-in care for transportation. Self-driving cars would help them get out of the house on their own.
In the ride service industry, consumers may feel safer in driverless vehicles than in traditional ones. They aren’t putting themselves in the hands of a stranger or worrying about their mental state. Best of all, there’s no risk of getting roped into a conversation.
Cons Of Self Driving Cars
There are undeniable problems in autonomous vehicles that must be addressed. One of the most prominent topics is these vehicles’ lack of human ethics and morals. In an emergency, an autonomous car may make a choice based on its training that increases damage to humans.
A less conceptual downside of self-driving cars is the loss of jobs in the ride-sharing industry. Every unmanned vehicle takes away a job. While the job may create other positions for programmers or risk assessment officers, the number of driver job openings will decrease. This shift disproportionately affects more unskilled workers.
The weather is yet another obstacle to the broad adoption of self-driving cars. The technology’s reliance on laser and camera sensors means road markings and signs can be hidden by snow or a curtain of rain.
The problem is the same for more rugged areas of the US. Many rural roads don’t have painted lines separating lanes or have one road for driving in both directions. Manufacturers rigorously test on private tracks, but the simulations typically expect a minimum road quality.
Conclusion
Self-driving cars have the potential to change our roads as we know them. The technology has a high potential to increase safety, convenience, and accessibility for countless people. However, it hasn’t reached the level of sophistication necessary to earn widespread trust.
The American people are hesitant to adopt autonomous vehicles despite the many benefits. Recent events like the Tesla recall make putting their loved ones in the backseat hard for many.
Despite this mistrust, driverless cars have succeeded in taxi services like Waymo. The difference is that these vehicles are restricted to cityscapes with ample data to learn from rather than back-alley roads or unknown suburbia.
Ultimately, the constant vigilance and reaction speeds of self-driving cars outpace the capabilities of human drivers. R&D is going strong, and we’re likely to see a massive spike in these vehicles on our roads within the next decade.
