Self-Driving Vehicle Safety: What You Need to Know

With AI, machine learning, and automation advancing at break-neck speed, many industries have sought to develop this technology to improve safety and efficiency. Self-driving cars, autonomous vehicles (AVs), and semi-autonomous vehicles are examples of how this technology can advance human initiatives and goals. However, self-driving vehicles still need further advancement — both to the technology itself and to the laws surrounding the new technology — for people to see their full potential. 

What Is a Self-Driving Car?

A self-driving car is just what the name implies — it’s a car that can drive itself with the help of technology. Self-driving cars are also used interchangeably with terms like semi-autonomous and autonomous vehicles. However, there is a difference between the two. Semi-autonomous vehicles often have the option to drive in auto-pilot mode, ideal in circumstances when drivers are cruising on a highway. Unlike autonomous vehicles, drivers in semi-autonomous vehicles must always be focused on the road to prevent an accident. 

Semi-autonomous and autonomous vehicles (AVs) are often placed on a scale based on their autonomous functions at the following levels: 

• Level 1: The driver undertakes most of the responsibilities, other than some assistance with steering, acceleration, and deceleration. Drivers must pay attention to all of their surroundings at all times.
• Level 2: The driver must provide feedback and pay attention to the car controls, steering, acceleration, and deceleration.
• Level 3: The driver must oversee the car as it undertakes all driving responsibilities.
• Level 4: A vehicle performs all driving responsibilities but cannot complete tasks in unsafe weather conditions, or untrained areas.
• Level 5: A vehicle performs all driving duties, regardless of the task.

According to the National Highway Traffic Safety Administration (NHTSA), full automation in cars, or cars at a level 5 of automation, is defined by the ability for “drivers” to be passengers in their cars instead. They don’t need to be engaged in planning the route or paying diligent attention to those around them — no matter the conditions. However, the NHTSA and other regulatory boards have yet to approve these vehicles, as most self-driving cars operate at a level 3 or 4 at maximum.

Key Technologies Behind Autonomous Vehicles

Semi-autonomous and autonomous vehicles use the following technologies:

• Sensor technology: Using radar, light detection and radiation (LIDAR), and cameras, a vehicle can recognize other vehicles, pedestrians, and other physical features outside the car. 
• Artificial intelligence: AI technology like deep learning uses data analysis and motion detection to understand surrounding traffic and traffic signals to follow through on a planned path. 
• Network infrastructure: Many vehicles use 5G technology and Wi-Fi to communicate with surrounding infrastructure and other vehicles to understand surrounding speed limits, signs, and traffic. The goal here is to prevent collisions by communicating with other technology.

This technology aims to create vehicles that respond quickly to stimuli so the driver doesn’t have to.

 Evolution and History

Manufacturers didn’t create autonomous vehicles overnight. In fact, this technology’s roots go back to the mid-20th century. Here’s a timeline of the technology.

• 1948-1960: Cruise Control enables drivers to maintain speed without using the gas pedal. Meanwhile, anti-lock brakes prevent the car brakes from freezing if the driver suddenly slams on the pedal to avoid a collision. 
• 1999: Adaptive cruise control, which aims to maintain cruise control by estimating the behavior of vehicles around it, is introduced.
• 2000s: Cars started featuring blind spot monitoring and forward collision warning to prevent collisions.
• 2013: Car manufacturers introduced automatic emergency braking to avoid oncoming collisions.
• 2014: Although review backup cameras long predate this event, Congress required their adoption in all cars in 2014.
• 2015: Using the same technology as the U.S. Department of Defense, Tesla introduces the autopilot feature, allowing drivers to drive hands-free on highways and freeways.
• 2017-2019: Manufacturers adopt lane departure warnings and lane-centering assistance to prevent side collisions.
• 2021  Cruise develops an autonomous vehicle that can operate in the pre-trained area of San Francisco.
• 2022: Self-driving services, such as Waymo, are offered to people over 18 in select cities.

From here, government agencies and private institutions aim to create fully autonomous cars that are safe for the public to buy.

Who Makes Driverless Cars?

Many car and vehicle manufacturers have released cars with some form of technology that categorizes them as semi-autonomous or at Level 1 of autonomy. These car manufacturers have mostly developed adaptive cruise control and lane assistance technology, which have largely become the norm among modern cars.

However, a few car manufacturers are aiming to gain full autonomy, including:

• Tesla: Tesla was one of the first companies to sell a car with an autopilot mode. Since then, it has released generations of cars with more features that test the boundaries of semi-autonomous and autonomous vehicle categories. As of 2024, its full self-driving system is currently in beta testing.  
• Waymo: Since 2022, Waymo — whose parent company is Alphabet — has offered self-driving cars as a service for people in cities like San Francisco, Austin, and Phoenix. 
• Cruise: As a subsidiary of General Motors, Cruise has had plenty of funds to develop a fully autonomous vehicle, which they released in 2021. However, General Motors postponed production in 2023 due to a fatal accident with a pedestrian and other incidents and has yet to announce plans for the company as they await an external investigation.
• Ford: Partnering with Volkswagen in 2021, Ford released Argo AI as a self-driving taxi service. However, they shut down the project in 2022 due to the uncertainty of ride-sharing projects’ profitability. Instead, they announced their development of a new generation of vehicles in 2023, hoping to sell them to the commercial public rather than as a service in the near future.

This, of course, is not an exhaustive list. Many other companies are dedicating budgets toward the development of AVs, but this list simply covers companies that have been forerunners in that initiative.

Current Autonomous Vehicle Services

Currently, the technology behind most autonomous vehicles is rated at level 4, meaning it can’t operate in an area it hasn’t been trained in or in unsafe weather conditions. However, as a service in a small area, passengers can enjoy the luxury of an AV vehicle. 

For the most part, existing ride-sharing apps like Uber and Lyft have partnered with the leading manufacturers of self-driving cars, such as Cruise and Waymo. They offer these services in Phoenix, Austin, and San Francisco. As of March 2024, Waymo announced its plan to extend operations to Los Angeles, as well.

In Las Vegas and Foster City, Zoox’s self-driving taxi service has remained profitable since it first released its shuttles in 2023. Vehicles released by this Amazon-funded company can only reach up to 45 mph. The company also isn’t large enough to build data for its cars to operate in other cities or further its manufacturing as of 2024. 

Are Driverless Cars Safer Than Human Drivers?

Although manufacturers sought out self-driving cars to make roads safer, there are various arguments about whether this goal has been achieved.

Safety Records and Statistics

In 2016, the first fatal accident of a semi-autonomous vehicle occurred in a Tesla Model S while it was on autopilot mode. Afterward, NHTSA published a report investigating the accident, claiming that the hardware and software were not to blame, but rather the driver’s inattention. This ultimately contradicts the goal of achieving a fully autonomous vehicle. If drivers must always pay attention, it isn’t fully autonomous even if the vehicle conducts all the driving tasks. 

Since fully autonomous vehicles are all in beta testing, few reports can comprehensively say whether or not autonomous vehicles are safer than their traditional counterparts. On the one hand, an NHSTA report found that semi-autonomous vehicles can reduce roadside accidents by 40% with features like lane assist. If you look at self-driving cars as a whole, though, they are twice as likely to experience an accident as traditional vehicles. 

Despite these statistics, many proponents of AVs argue that they have the potential to reduce fatalities with the help of lane assist, blind spot monitoring, and other collision prevention technology. 

Ongoing Safety Concerns and Criticisms

There are plenty of criticisms of self-driving cars, and most of them revolve around the fact that any current iterations or tests of fully AVs are incompatible with complex roadside circumstances, according to a comprehensive 2023 University of Texas at Arlington analysis. 

Propelled by machine learning, though, these cars are constantly learning how to make a car’s occupants safer. That said, it may not provide the same thought for others on the road, particularly pedestrians. Current generations of AVs have difficulty distinguishing between a pedestrian about to cross the street and a pedestrian walking on the sidewalk or looking through a store window.

Meanwhile, emergency services have complained about other risks self-driving cars pose in impeding their route to accidents. In these circumstances, every second matters, making any delay from AVs a risk.

What Are the Benefits of Self-Driving Cars?

Despite the concerns about self-driving cars, they show a promising future, and there is good reason for that.

Fewer Accidents Caused by Human Error

Many proponents of AVs argue that the technology eliminates the cause behind most traditional vehicular accidents: human error. In an NHSTA report, 94% of accidents occurred because of distracted driving, drunk driving, or bad decisions. AVs could reduce this statistic by using technology to enhance awareness of the road around them and remove any emotion that may come from an oncoming collision. 

Efficiency and Traffic Management

Accidents become more frequent as traffic congestion increases. AVs can avoid this outcome altogether by evening out the flow of traffic. AVs do this by reducing “phantom traffic jams” when drivers speed up to the car in front of them instead of leaving an even distance behind. 

Simultaneously, cars can be much more ecologically efficient when they follow this practice, as pressing on the gas and the brake pedal is more likely to increase fuel consumption and greenhouse gas emissions. In a UC Berkley study, AVs who used this practice were able to reduce their emissions by 42% on a practice course with other human-controlled vehicles.

Accessibility

Most notably, many AV experts believe they can make driving more accessible for people who may not have been able to drive in a traditional vehicle and for the general public. By offering more modes of transportation, people have increased mobility, which can be especially difficult in cities where walkability is an issue. 

For people with disabilities, traditional vehicles are typically not a viable or affordable option. For example, someone who uses a wheelchair may not be able to afford modifications that a traditional vehicle requires for them to use. In addition, someone with a visual impairment may not even be able to drive entirely, while an AV could solve a significant transportation problem. 

In addition, regulatory agencies often bar older adults from driving once their cognition starts to decline, which can severely limit their freedom to go wherever they want. If AVs become fully autonomous, these drivers don’t have to worry about safely reaching their destination.

What Are the Drawbacks of Autonomous Vehicles?

As mentioned earlier, companies like Cruise or Waymo aren’t large enough to start training their AVs elsewhere other than a few select cities. It would take a lot of data and time to collect and analyze that data for AVs to be viable on all roads across the country — assuming that it will be a future outcome. Currently, most Americans — about 44% — hesitate about this future.

Let’s look at some of the reasons behind this hesitation.

Technological Reliability

Just as a computer glitches now and then, an AV has the potential to do the same, but with dangerous impacts. Even semi-autonomous vehicles, such as Tesla models with the autopilot feature, have misinterpreted obstacles and crashed into other vehicles or pedestrians.

Just as concerning, if an AV is connected to Wi-Fi, hackers can potentially infiltrate the system and take over the car — even disabling the brakes, as one 2015 accident illustrated.

As such, companies like Waymo have announced that their cars can disconnect from the internet to prevent these situations. They can do this by storing data on local computers in the car rather than a cloud system they communicate with. 

Ethical and Moral Dilemmas

The AV also complicates an ongoing question philosophers have had for centuries: if given the choice to save one individual versus a group of people, who would you save? Although this situation, known as “The Trolley Problem,” is always given a theoretical setting, it could very well be tangible for AVs. 
For example, an AV may avoid an individual bicyclist and swerve into oncoming traffic, potentially injuring countless others. AV programmers argue that this is a fruitless argument, as they train AVs to prioritize the standard of care for their occupants and traffic laws, which often prevent circumstances like this. However, many engineers are still uncertain how an AV would act in the Trolley Problem, leading to further debate over the ethics of giving control to a “robot.”

Economic and Employment Impacts

The biggest worry with integrating automation is how it’ll affect employment and, in a broader scope, the economy. According to Andreas Tschiesner, a senior partner at McKinsey who specializes in the Automotive and Advanced industries sector, AVs have the potential to replace shuttle drivers, taxi drivers, and other gig drivers. In addition, the more convenient AVs become, the fewer people will turn to public transportation, providing less funding to civil servants within this sector.

Like automation, AVs won’t create an entirely robotic landscape within the transportation industry. The advent simply displaces these employees, requiring more employment in control traffic centers and fleet management to prevent road congestion. 

Challenges in Mixed-Traffic Environments

The transition period is proving to be the most challenging for experts in AV development. Not everyone drives an AV; moreover, not everyone drives as an avenue of transportation. This presents difficulties as AVs advance their algorithms to learn more about traffic habits and prevent accidents. Until AVs are normalized, accidents may continue to occur, discouraging the public. Thus, the cycle of AV development will stagnate.    

Do Self-Driving Cars Have Fewer Accidents?

The short answer is yes, self-driving cars have fewer accidents if you look at the primary tallies of accidents on the road. However, this conclusion is unsteady, considering that manufacturers are still testing AVs. In addition, the ratio between AV and traditional vehicles is highly disproportionate, leading to unbalanced totals. In addition, people have yet to use AVs in inclement weather conditions such as fog or snow, which are often responsible for traditional vehicle accidents. 

Let’s examine existing research that can help answer this question more comprehensively.

Existing Accident Research

In the 2023 University of Texas Arlington analysis of AV accidents mentioned earlier, researchers evaluated more than 50,000 accidents of traditional and autonomous vehicles, the types of crashes, damage, and causes of the incident. Compared to traditional vehicles, most AV accidents happen in complex environments such as intersections, while traditional vehicle accidents occur on the freeway or highway. 

The data pointed to the fact that AVs will brake suddenly because of an unexpected circumstance — such as a pedestrian walking into traffic — causing the vehicle behind them to rear-end them. These accidents often occur at low speeds as opposed to traditional vehicle accidents, leading to minimal injury. Because most traditional vehicle accidents happen at high speed on the highway, fatalities are more common. 

Although we shouldn’t ignore the frequency of AV accidents, the data provided illustrates that AVs decrease the frequency of fatal accidents compared to traditional vehicles.

What Happens When You Are in an Accident with a Driverless Car?

Even if you’re less likely to get injured while in a driverless car, accidents still happen, especially to others on the road around you, while these circumstances are still unique, they are becoming more common, and it’s best to be prepared and know how to proceed.

Immediate Steps and Reporting

Like in any other traditional car accident, you must first check for everyone’s safety. Then, you can call emergency services for help or instruct someone else to do so. Afterward, exchange insurance and contact information with everyone involved. Don’t forget to document any damage that happened, such as personal injuries or property damage. You’ll need this when talking to your insurance, the authorities, and lawyers later. Finally, you can call your insurance agent to proceed to future steps.

Insurance and Liability Issues

Since AV accidents are relatively new, your insurance will most likely handle any incident with an AV similar to a traditional car accident, or accidents with ride-share companies. Essentially, drivers in AVs will be held responsible for any damages if they aren’t paying attention while in the car. However, this can get complicated, especially considering some of the faulty habits some AVs have had. In these cases, it may be best to proceed with legal action with the AV manufacturer. 

Legal Recourse and Compensation

Since AVs are a relatively new adoption on the road, consulting a legal professional at Heidari Law Group may be best, no matter the outcome of the accident. They’ll know how to proceed in any car accident, and since their practice areas do feature AV services, they may know more options on how to proceed. 
Suppose you have a personal injury from an accident with an AV vehicle. You may be liable for any medical expenses from either other parties involved in the accident or even the AV manufacturer itself. As more cases of AV accidents come forth, legislators and other public policymakers can use it as a legal precedent for future AV legislation and regulation.

Legal Ramifications of Self-Driving Cars

AV development is moving faster than government officials can keep up. At first, many were in full support of AV development. Still, there are many hoops that manufacturers and policymakers need to jump through before these vehicles become standard in the public.

Current Regulatory Landscape

As of 2024, the NHSTA has yet to approve a fully autonomous vehicle. However, in September 2019, they published thorough guidelines for AV safety protocol. The guidelines also state that companies and other entities don’t need to wait for federal approval to test AVs; they must gain approval from state legislation. Since then, many states have created legislation or gained executive action to enable AV testing and improvement. 

In 2021, NHSTA passed an act that required all AV manufacturers and owners to report any crashes of AVs over level 2 automation to the agency as a way to compile data and better enact safety requirements on AVs.

Liability and Insurance Reforms

As of 2024, the U.S. government has yet to make official changes to liability laws for AVs. Like traditional vehicles, “drivers” in AVs assume the responsibility for any damages they’re responsible for, as they should still maintain awareness while in the vehicle. That said, NHSTA has been vocal about the need for liability and insurance reform before manufacturers release fully autonomous vehicles to the public. 

Some experts have argued that existing tort laws can help guide the legal framework so legislators don’t waste time creating entirely new laws for AVs. AV manufacturers will be held accountable for accidents as long as significant proof is available. If this comes to pass, AV servicing companies may adjust pricing to account for any insurance they may have to pay for in these cases. 

Future Legal and Ethical Considerations

Moreover, policymakers should also consider the future impacts AVs could have on public transportation and traffic congestion. If hiring a self-driving taxi is convenient, people may be less likely to share a subway with dozens of others or walk to work. In this case, traffic on the road may become busier, and AVs will have to work harder to even out the traffic flow. 

Road infrastructure may also need to change to minimize intersection interactions that confuse AVs and cause car accidents. Before this can happen, though, AV companies need regulation so they can’t neglect their responsibility for any personal injuries or accidents due to their vehicles. 

Many experts believe that AVs will become standard across the U.S. by 2030. Time will tell if this estimation proves true, especially considering all the roadblocks self-driving cars have had. Until then, stay vigilant on the road and seek legal help when needed.

FAQs on Self-Driving Vehicle Safety

What Is a Self-Driving Car?

A self-driving car is a vehicle that uses artificial intelligence, sensors, and cameras to operate a vehicle without the supervision or assistance of a driver. Theoretically, they should be able to drive anywhere in any condition. However, manufacturers are still testing autonomous vehicles (AVs) to be fully capable of driving in any setting, not a select location in various weather conditions. 

Are Self-Driving Cars Available?

As of 2024, no fully autonomous car has been approved for sale to the general public. However, certain companies have provided self-driving services in certain trained locations, such as Las Vegas, San Francisco, Los Angeles, and Phoenix. Cars with semi-autonomous features, such as autopilot mode or automatic emergency braking, are also available for sale. 

How Safe Are Self-Driving Cars?

There is not enough testing to conclusively say, but self-driving cars have been more prone to accidents without the supervision of a driver. Still, they have fewer fatalities than traditional vehicles. 

What Are the Benefits of Self-Driving Cars?

Self-driving cars reduce accidents caused by human error and make driving more accessible to the public, people with disabilities, and older generations. They can also improve traffic flow by keeping a steady pace with the cars around them. 

What Are the Challenges of Self-Driving Cars?

Current AVs aren’t reliable in nuanced environments, particularly in intersections. Current models have been prone to hacking and faulty technology. Many have pondered the ethical and legal implications of AVs on the road, as AVs mainly prioritize their occupants over others on the road. They may encounter issues when deciding between colliding with an individual or a group of vehicles or people. AVs would also remove the need for public and private transportation employees and drivers. Legislators would also need to update liability laws to account for AVs.

What Do I Do After a Self-Driving Car Accident?

As in any other car accident, make sure everyone is safe, call the authorities, exchange contact and insurance information with everyone involved, document any damage involved, and call your insurance provider and legal team.

Who Is Liable in a Self-Driving Car Accident?

As of 2024, drivers in self-driving cars assume any liability if they are responsible for an accident, as most AVs still require their attention. However, more policymakers and government organizations are pushing for updated changes to liability in these cases. Some experts say that legal precedent from tort cases could hold manufacturers liable due to faulty programming.