Self-driving cars, also known as autonomous vehicles, are becoming more common and may soon be the majority of vehicles on the road. Driverless cars and trucks are seen on the streets more frequently as companies pursue testing to assure that the vehicles can navigate the roads without mishaps. Fully driverless vehicles outside of the testing phase are expected to arrive by 2020. Let’s look at self-driving cars, how they work, how accidents will be addressed, insurance and liability. Without a doubt, what was thought of as futuristic a few years ago is here.

In this article:

• Self-Driving Cars Legal Representation
• How Do Self-Driving Cars Operate?
• Incoming Information on Self-Driving Cars
• How Do Self-Driving Cars Learn?
• Where is the Technology Today?
• When Will Self-Driving Cars Be Available?
• Will Fleet Vehicles Be the First?
• Advantages of Self-Driving Cars
• Getting Drunk Drivers Off the Road
• Making the Highways Smart
• The Work is in the Cloud
• Changing the Perception of Mobility
• Driverless Trucks
• Future of the Trucking Industry
• Self-Driving Delivery Vehicles
• Changes in the Insurance Industry
• Issues About Ethics
• Statistics on Self-Driving Accidents
• Rules of Liability
• Pros and Cons of Self-Driving Cars

Overall, laws follow technological advancements. This is a new era in transportation, much like the change from horse and buggies to automotive transportation. So as driverless cars become commonplace, laws will be modified to accommodate this technology. During this transition, emphasis will be placed on becoming as familiar with the legal stipulations as we are in other areas. It places a demand for lawyers to learn how legal cases dealing with self-driving cars are handled. Meeting this requirement is an obligation legal professionals share. Our law firm will be at the forefront of this new technology from a legal standpoint. We want our clients to understand the intricacies of this new type of transportation as we move together into an exciting and new future of transportation.

Self-driving cars are run by an intricate system comprised of software, cameras, and sensors. Sensors and cameras enable the vehicle to “see” and navigate the terrain. Vehicles are still in the testing phase, but soon, level 4 vehicles will be available without the need for human observation or intervention in some situations. According to experts, there are five levels a vehicle can attain, each with varying ability:

• Level 0: Humans control all functions.
• Level 1: Some control different driving tasks such as braking or cruise control without human intervention.
• Level 2: Some autonomous driving tasks such as steering or acceleration can be done simultaneously, but human intervention is still involved when needed.
• Level 3: Under particular conditions, the vehicle can perform all safety-oriented functions, but a driver must intercede when alerted.
• Level 4: Autonomous operation is fully operational in specific situations. Others require the assistance of a human driver.
• Level 5: The vehicle is completely autonomous without the need for human intervention. This technology is under development in Chrysler Pacifica models at the current time by Waymo.

Self-driving cars use a series of sensors to build a map of the area through which the vehicle is passing. The maps are then used to instruct the vehicle’s actuators that control everything from steering to braking. The vehicle’s “brain” learns to differentiate between images using artificial intelligence. Algorithms within the software enable the self-driving car to apply the rules of the road and traffic laws.

Self-driving cars can learn through the use of artificial intelligence (AI). This forgoes the usual hard algorithm of the “if, then” rules used by computers. For instance, teaching the vehicle to recognize an object is fed a long list of objects. If the vehicle is shown an image of a ball, it is expected to identify the ball from the list of objects it has in its memory. At the outset, the vehicle will more than likely choose the wrong object. But, over time, its accuracy will improve maximally. The use of AI is an important part of how the self-driving car works.

The use of multiple layers of technology is already being used in prototypes of fully self-driving cars. These include:

• AI (artificial intelligence)
• Cameras
• Connectivity
• Data management
• High-efficiency GPS
• Internet of Things (IoT)
• LIDAR (light detection and ranging)
• Radar
• Sensors
• The cloud

Level 1 features, such as advanced cruise control that can slow down and accelerate on its own, lane control, and automatic parking and braking, are already available in some vehicles. Some high-end vehicles have autonomous steering and braking, albeit over short distances. Level 3 and 4 vehicles are still being tested and require human input, so steering wheels and brake pedals are still present in these vehicles. Level 5 vehicles will lack a steering wheel and brakes.

While the introduction of mass-produced commercial and privately-owned self-driving cars has not really begun, it should begin in earnest within a few years. General Motors has said that the company has a vehicle that is ready to be mass-produced. Still, the company is waiting for testing to be completed and legislation in place before going forward. The general estimate is that by 2020 autonomous cars will be taking to the roads en masse. By 2040 it is thought that 95 percent of all new vehicles sold will be autonomous. This means that over 96-million driverless vehicles will be on the roads, worth about $3.6 trillion. Tech-oriented companies, including Tesla and Google, are leading the way. Both have already mapped out millions of ground miles. A Google Alphabet subsidiary recently announced that it had surpassed the three million mile mark. Other carmakers are moving into position, and Ford expects commercial vehicles to be on the road by 2021.

Fleet vehicles will undoubtedly be the first on the road. This is due to several things, such as the ability to monitor the vehicles. Fleet vehicles, as opposed to those sold to private owners, are easier to monitor. By having this access, fleet owners can have repair teams on alert to deal with any problems. In addition, the likelihood that fleet owners will maintain and service their vehicles is higher. Since taxis or other fleet vehicles operate in cities, the speed rate will be lower than highways. This may be beneficial in the early days of self-driving cars.

The World Health Organization estimates that over 1.3 million lives are lost due to motor vehicle fatalities globally. In the United States, approximately 30,000 lives are lost, with roughly 10,000 due to drunk driving. Since both distraction, fatigue, and human error, as well as drunk driving, are not a part of a self-driving car’s “behavior,” many lives will be saved. This increased safety will lead to savings of over $234 billion in medical costs, lost time from work, and other expenses.

Self-driving vehicles may seem like an idea that will never come to pass, but several car companies have already placed autonomous cars on our roadways. Check out the video below:

While mentioned above as an advantage of using driverless technology, this category requires additional mention. The use of driverless taxis and privately owned vehicles and vehicle-sharing operations will stop serious injuries and deaths caused by drunk driving on the roads. A major lobbying group, trying to ensure that legislation will move faster to get driverless vehicles on the road, counts tech companies and carmakers among its members. A new addition to this group is the Wine & Spirits Wholesalers of America. By dealing with impaired driving, an entire industry that relies on imbibing spirits will benefit.

Self-driving cars driven by software will make the transition to smart streets possible. For instance, it will be possible to integrate traffic signs and lights and merge lanes more efficiently, eliminating traffic congestion. The ability of self-driving cars to interact with each other and the infrastructure will make traffic less of a problem. The need for parking spaces will also be reduced as autonomous vehicles can drop a family member off at work, deliver children to school and take them to baseball and soccer practice without taking up parking spaces.

The use of cameras and sensors makes the amount of data received astronomical. The data must be used in real-time, which means the capability to process it must be available instantaneously. The effectiveness of data processing will enable the safe driving of the autonomous vehicle. Other data, such as updates to the system and changes in the learning modules, will need to be current. Only a fully functional and efficient cloud-based system can be counted on to handle this.

The interaction between vehicles will also rely on the cloud. This will enable vehicles to share knowledge of changes in road conditions and even detours. Destination location could be available in the cloud or onboard the vehicle. The determining factor would be the availability of geographic information from one area to another. As more roads are mapped, and “smart” capabilities are applied, most data will be in the cloud.

The concept of travel will change with the advent of driverless vehicles. The concept of car ownership will transition to serviced mobility. This frees up driver time when commuting from one location to another. Suddenly, the burden of travel will switch to one where a commuter/traveler can relax, read a book or catch up on work. The time-saving aspects of this will be invaluable to people who never seem to have enough time to get things done.

Long-haul driverless trucks are poised to become a big part of the trucking industry. This is because highway travel is easier to navigate than more intricate urban travel. Overall, $700 billion worth of cargo is moved every year by trucks. New legislation that restricts the number of hours a trucker can work and the catastrophic accidents when drivers are pushed beyond their limits can affect getting cargo to its end destination quickly enough to satisfy demand. Autonomous trucks with lower overhead and faster turnaround times will make driverless cargo transport welcome within the industry.

The trucking industry has a problem finding enough drivers to handle shipping demand, particularly among younger drivers. The turnover rate for truckers is approaching 95 percent, according to the American Trucking Association. Estimates say a driver shortage of 50,000 will exist soon. Such factors point to the need for an overhaul, and autonomous shipping might be the answer.

In 2016, a report issued by the White House said of the 3.5 million truckers, 80 percent might lose their jobs due to the use of autonomous trucks. Other groups have countered that by saying that driverless trucks, at least in the beginning, would still require the assistance of a trucker. However, they would be free to rest while the truck drove, getting cargo to its destination quicker.

Truck manufacturers, including Waymo, Uber, GM, and Toyota, among others, and foreign manufacturers such as Volvo, are leading the way into driverless truck automation. Volvo already has an autonomous truck operating in Sweden’s Kristineberg Mine, which will be instrumental in augmenting safety and operational flow.

Waymo, formerly part of Google’s driverless car project, has announced that its autonomous trucks would begin hauling freight from Google’s data centers in Georgia. Due to Atlanta’s status as one of the country’s largest cargo hubs, the company feels it is the perfect place to try out its autonomous truck deliveries. Its Class 8 tractor-trailers were first tested in Arizona in 2017.

In March 2018, Uber announced that it would be hauling commercial freight in Arizona. The Uber system is a three-stage process for delivery and uses truck drivers at the beginning and end of the journey with automated trucks hauling cargo long distances. Truck drivers are better equipped to handle the complexities of docking, loading, and unloading trailers and busy trucking facility yards, as well as navigating through heavy traffic on non-highway roads.

Initially, Uber is confining its activity to Arizona due to easy self-driving car regulations in the state and other legal reasons. Trucks are making the 300-mile trip between Topock in the western part of the state and Sanders in the east. The company is planning to expand the territory to multiple states in the future, using its cargo-hauling network Uber Freight. This model allows shippers to find cargo haulers and truckers much the way the Uber taxi service does.

Starsky Robotics is another player in the autonomous truck market. In February 2018, it became the first to haul cargo in a fully autonomous cargo model on roads in Florida. Like Arizona, Florida’s laws about driverless vehicles are lax, allowing for an easier transition from semi-autonomous to fully autonomous models. Unlike Uber, the vehicles do final destination delivery by keeping the driver out of the truck and monitoring the self-driving car from an office. Using a call-center-like setup, drivers can monitor the truck from pickup to the highway to the delivery point. It is used in truck yards and freight corridors, hauling such cargo as milk and tile weighing 40,000 pounds. This, the company feels, will help the nation’s truck driver shortage. While Starsky Robotics is limited to operating in Florida now, it plans to move into multiple states.

Embark, a company that counts Ryder Trucks, and Electrolux, has completed the first coast-to-coast automated truck voyage with a driver onboard. The 2,400-mile trip stretched from Los Angeles to Jacksonville, Florida. One difference with the Embark model is that it does not rely on long-distance mapping but rather uses AI to allow the learning process. The company’s vehicles rely on sensors for real-time data only.

The video shows a fully self-driving truck being tested in operations deep underground in the Kristineberg Mine.

Delivery trucks are first in line to become autonomous. Several companies are working to transition prototypes to the roadway. One tech company, Nvidia, is partnering with DHL to test a fleet of self-driving cars. Nvidia uses artificial intelligence to get the van to its destination.

Other contenders in this market are Nuro, a company led by ex-Google engineers. Their pod-like vehicles without seats will be used to deliver items such as groceries or online purchases to a customer’s doorstep or curbside in front of their home or office. An alert would be issued letting the customer know the package(s) have arrived and provided a code that the person can unlock a compartment in the vehicle.

Toyota is forming partnerships with Pizza Hut and Amazon to develop a self-driving delivery vehicle called the e-Palette. Estimates from management are reporting it will be available around 2020. Since the e-Palette has a 13×23-foot length, it can hold small stores that will come to you. Likewise, Ford is testing a specially-built Fusion to deliver Domino’s pizza in Michigan. Customers will go to the car when it arrives and extract their pizza from an ATM-like structure attached to a warming oven in the vehicle. The big question is whether Domino’s customers will be comfortable with this sort of delivery.

The insurance industry will face a changing landscape with diminishing auto premiums due to a decrease in accidents. Decreased premiums will also be lowered by reducing individual car ownership and reliance on autonomous fleet vehicles. A decrease in revenue will most likely occur after an estimated bump of $81 billion by 2025. However, by 2035, decreases in revenue will become apparent.

The insurance industry will direct its attention to three areas that can boost its overall revenue such as:

• Cybersecurity: Expected to generate about $12 billion, this area will focus on misused data, hacking, and ransomware, preventing user access to their system.
• Product liability: This represents a costly area since chips and the numerous sensors on self-driving cars are expensive to repair or replace. The failure of software is another issue that may present massive liability problems.
• Insurance for Infrastructure Issues: Since the government has its own insurance structure, this may be the least lucrative area for insurers. However, problems in the cloud or other safeguards will require protection.

Unless the insurance industry adapts quickly, their losses will probably total $15 billion by 2025, $25 billion by 2035, and $50 billion by 2050, according to reports in the Insurance Journal.

David Gerdes, a Stanford professor in mechanical engineering, questions how safety concerns, which predominate, will be balanced by ethical concerns. The preeminent focus of automated vehicles is to keep the occupants safe. There are times when a choice must be made to preserve the safety of someone outside the vehicle, such as a small child who runs into the driverless vehicle’s path. What if avoiding hitting the child results in possible injury to the occupants of the driverless vehicle?

Human drivers may choose to protect the child and take the risk of injury upon themselves. The question is, what would the self-driving car do? Since it is programmed to stop before hitting an object or person, the vehicle will stop. Since the ability to stop is based on the recognition of an object ahead, a human driver would most likely not stop in the time since stopping requires enough time to brake and force the brakes to overcome the vehicle’s velocity. The driverless vehicle could accomplish this smoothly.

However, if the child simply appears out of nowhere, as they so often do, the question engineers have to solve is getting the self-driving car to stop at a moment’s notice. If this cannot be done, ethics experts debate whether the car will simply choose to keep the occupants safe as it is programmed to do.

It is somewhat early to provide statistics on car accident incidence. However, it is possible to offer statistics using autonomous testing vehicles. Starting in 2014, the California Department of Motor Vehicles (DMV) received 59 instances of collisions involving a self-driving car.

The accidents were overwhelmingly caused by drivers in non-autonomous vehicles. Most were accidents where the vehicles rear-ended self-driving cars. Others were caused when another vehicle hit the self-driving car during lane changes or when the self-driving car was turning left. An example of a crash is as follows:

On December 7, 2017, a GM Cruise self-driving car in San Francisco was in the center lane of three adjacent one-way lanes when it attempted to move into the left lane after seeing a space between two vehicles already in that lane. As this was attempted, the lead vehicle slowed down, and the Cruise returned to the center lane of the three lanes. Simultaneously with that, a motorcycle, which had been lane-splitting between a vehicle in the right-most lane and the center lane, moved into the center lane, hitting the Cruise and falling over. The speed of the Cruise when the accident happened was 12 mph, and the motorcycle’s speed was 17 mph. After hitting the pavement, the biker moved his vehicle to an area on the side of the roadway, and both he and the Cruise exchanged insurance information. After examining the accident, it was determined the biker was at fault for trying to pass a vehicle on his right-hand side without having clearance to do so. Self-driving cars are required to carry insurance information on their computer systems.

The type of liability a self-driving car has after a car accident depends on whether or not the driverless car or truck caused the accident. If it was not at fault in the accident, then the usual rules of liability apply. Since non-autonomous vehicles will share the road with autonomous vehicles, accidents will still happen. As mentioned earlier, since driver error causes most accidents followed by reckless driving or drunk driving, the number of accidents will decline.

However, if the self-driving car did cause the crash, the rules change. Since a driver is not present, common reasons such as driver error, drunk, or otherwise reckless behavior are no longer in the mix. The most likely reason a self-driving car would cause an accident would be due to a flaw in its software, failed sensors, or infrastructure problems. The latter would most likely be the fault of a city, state, or federal entity, and a claim could be filed against the at-fault party.

If the software or sensors are to blame, then a lawsuit against the vehicle’s designers, its manufacturer, or the supplier of parts would be to blame. The focus of a lawsuit would switch to auto product liability. It is not necessary to prove negligence in a product liability case in California due to the doctrine of strict liability at use in the state. Still, it is necessary to prove that the flaw did cause the accident. A product liability lawyer would go forward with a case upon the responsible party or parties.

The pros of driverless cars are:

• They are safer. Eighty-one percent of accidents happen due to driver error. Driverless cars rarely make a mistake.
• No distraction: self-driving cars are never distracted.
• Stopping is accurate: Since self-driving cars use precise measurements to stop, the chances of accidents are nil unless a malfunction occurs.
• Lives will be saved.
• Self-driving cars are time efficient: No more spending hours driving to and from work. The amount of time someone can save in a year is roughly 365 hours if it takes an hour to get to and from work. That is roughly 15 days.
• Improve traffic: Since autonomous cars drive efficiently, they will decrease congestion and the use of fuel.
• Help the disabled and elderly get around: This is a major benefit of driverless vehicles. The old and disabled will no longer need to rely on others to do errands, increasing their sense of freedom.
• Speed limits may increase: Due to the safety of driverless vehicles, speed limits may be increased. This will save time and fuel.
• Drunk driving reduction: Since impaired individuals will no longer be behind the wheel, the chance of a drunk driving accident is mostly eliminated.
• Police will be free to work on serious crimes rather than accidents and reckless driving.
• Freedom from parking: A self-driving car would be able to take you to your destination and park six blocks away.
• DMV would no longer need to give out licenses and test drivers.

Cons of self-driving cars are:

• Cost: The first self-driving cars will probably be costly for personal use. Fleet vehicle use will work out better.
• Programs could be hacked: Hacking and malware threats will require vigilant oversight.
• Jobs will be eliminated: Jobs in transportation will be hurt, particularly taxi drivers and truck drivers. Industries that support truck drivers, such as truck stops, will go away since many autonomous trucks will be electric or hybrid, extending their range.
• Inclement weather may cause problems: Sensors and cameras may have interference from bad weather. Also, communication with the cloud may be affected.
• Interpreting directions: driverless cars would be unable, at least at this time, to interpret construction flags or worker hand signals to stop or proceed. Likewise, the vehicles would not understand a police officer telling motorists to proceed when traffic lights break down. Manual signals are not decoded by the algorithms embedded in the vehicle’s computer system.
• Deciphering the law: a vehicle may not be able to adapt to all state laws.