Autonomous Vehicles
Autonomous cars are no longer merely a concept from science fiction films, as they are increasingly beginning to operate in the real world. This type of vehicle is designed to take over the driver’s responsibilities, which means not only controlling speed and the course of travel, but also making decisions in dynamic road situations. Growing interest in this technology stems from the need to improve road safety, enhance traffic flow, and provide greater comfort for users. At the same time, the development of artificial intelligence has enabled machines to analyze their surroundings more quickly and accurately than humans, opening a new chapter in the history of motoring.
In recent years, many companies have been investing heavily in research into autonomy. This is driven both by growing competition and by the awareness that demographic changes, urbanization, and the increasing importance of mobility services will fuel the need for modern forms of transport. Autonomy is not only a technology but also a reorganization of how we will think about road traffic, vehicle ownership, and the role of drivers.
Table of contents
How Does Autonomy Work in Vehicles and What Components Make It Up
An autonomous car is, above all, a system of interconnected components that seeks to understand its surroundings and respond appropriately to changes. The key is data, which the vehicle must receive, process, and use to make decisions. Unlike humans, a car does not rely on a single type of information. Instead, it builds a multilayered map of reality by combining camera images, radar signals, lidar measurements, and GPS information.
An important part of the process is also data interpretation. Simply reading signals is not enough if the system cannot translate this information into real action. This is precisely why artificial intelligence plays such a crucial role. It enables the analysis of road situations, the prediction of other road users’ movements, and the making of decisions that, in traditional terms, were the domain of humans.
Autonomous Car Models
Tesla Model S
One of the best-known models with advanced autonomous features is the Tesla Model S (particularly the version equipped with the Full Self-Driving package). This electric sedan can recognize traffic lights, change lanes, and follow traffic in many road situations, although it still requires the driver's attention. Another example is the Tesla Model 3, a popular compact electric car that gains additional features supporting autonomous driving through software updates.
Mercedes-Benz EQS
Besides Tesla, it is worth paying attention to the Mercedes-Benz EQS, a luxury electric sedan equipped with the Drive Pilot system (approved for use under specific conditions in selected countries), which enables autonomous driving in traffic jams and on motorways at limited speeds. The BMW iX also offers advanced driver assistance (BMW Personal CoPilot), which can take over control of the vehicle in many scenarios, minimizing the driver’s effort.
Waymo One
In the segment of robotic taxis and shuttles, Waymo One has emerged—a vehicle based on Chrysler Pacifica Hybrid models (modified for fully autonomous driving) that, in selected U.S. cities, transports passengers without a driver in the front seat. Similar solutions are being tested by Cruise Origin, an electric vehicle without a traditional cockpit, designed for ride-sharing and autonomous urban mobility.
Artificial intelligence as the car’s brain
It is precisely artificial intelligence that makes it possible to call a vehicle autonomous at all. Algorithms learn from vast datasets that encompass various driving conditions, accident situations, driver reactions, and weather changes. Predictive models analyze the behavior of pedestrians, cyclists, and other cars, and then create the most likely picture of what will happen on the road. This allows the car to anticipate events rather than merely react to them as they occur.
Artificial intelligence is also responsible for risk assessment. When a sudden stimulus occurs, such as a child running into the road, the algorithm analyzes thousands of possible courses of action. It takes into account the distance, speed, and direction of the vehicle, the condition of the road surface, and potential hazards. All of this happens in a fraction of a second. This efficiency and ability to process information far exceed human capabilities, which is one of the arguments that autonomous vehicles can increase safety.
Assisted and Other Indirect Solutions
Between a classic car driven entirely by the driver and a fully autonomous vehicle, there is a broad range of intermediate solutions. These are the ones most commonly taking to the roads today, and in practice they are changing the way we drive, even if they still formally require a human to be present and responsible behind the wheel.
A good example is adaptive cruise control systems combined with lane keeping. In the Toyota Corolla and Toyota RAV4, as part of the Toyota Safety Sense package, the car can independently accelerate, brake, and gently correct its trajectory on the highway. The driver keeps their hands on the steering wheel, but their role is mainly limited to supervision. The Volkswagen Passat with the Travel Assist system works similarly, combining adaptive cruise control, traffic sign recognition, and lane centering.
In the premium class, such features are even more advanced. The Audi A6 and Audi Q7 offer traffic jam assist, allowing the car to move autonomously in heavy traffic, respond to vehicles stopping and starting, and maintain a safe distance. In the BMW 5 Series, the Driving Assistant Professional system can even suggest a lane change and perform it after confirmation by the driver, demonstrating how close these cars are to semi-autonomous driving.
A separate category consists of automated parking systems, which for many drivers are their first real contact with autonomy. Ford Kuga, Skoda Octavia, and Hyundai Tucson can park parallel or perpendicular on their own, taking control of the steering wheel and often also the accelerator and brake. BMW iX and the Mercedes-Benz S-Class go even further, making it possible to park remotely using a smartphone, without anyone sitting in the car.
It is also worth mentioning 360-degree camera systems and maneuvering assistance features, found, among others, in the Nissan Qashqai and Kia Sportage. Although these are not formally autonomous solutions, they significantly reduce the risk of error when driving in the city and in tight spaces. The car monitors its surroundings, warns of obstacles, and can automatically stop when it detects a risk of collision.
Autonomous Trucks
The development of autonomous trucks is following a different path than that of passenger cars and is, in many respects, more advanced today. This is due to their simpler operating environment, predictable routes, and the significant economic benefits that autonomy can bring to freight transport.
The most advanced deployments concern highway driving. Long, monotonous stretches of expressways are significantly easier to automate than urban traffic. Companies such as TuSimple, Aurora, Kodiak Robotics, and Plus have been testing trucks for several years that drive themselves on highways, maintain their lane, speed, and distance, while the driver acts as a supervisor or is not in the cab at all during closed-course tests.
In practice, this means a hybrid model. The autonomous truck takes over driving on routes between logistics centers, while maneuvers in cities, at ramps, and in loading yards are carried out by a human or a remote operator. This division significantly lowers the barrier to bringing the technology to market and makes it possible to avoid the most challenging road scenarios.
Commercial vehicle manufacturers are also investing heavily in autonomy. Volvo Trucks, Daimler Truck, and Scania are developing autonomous systems for both long-haul transport and operations in enclosed environments. Automation is progressing particularly rapidly in ports, mines, and logistics centers, where autonomous trucks travel along strictly defined routes and do not have to respond to unpredictable road users.
Legally, autonomous trucks often fall under the same regulations as passenger cars, but in practice they have an easier time. In many countries, tests without a driver in the cab are permitted, provided there is remote supervision and strict reporting. The United States is a leader in this area. In some states, autonomous trucks are already carrying out regular test transports on designated routes without a safety driver.
A key factor driving development is the labor market. The transport industry has been struggling for years with a shortage of drivers, rising costs, and working-time restrictions. Autonomy makes it possible to extend a vehicle’s operating time, increase delivery predictability, and reduce fuel consumption through smoother driving. For this reason, the economic pressure to implement autonomous solutions is significantly stronger in heavy-duty transport than in the case of private cars.
As a result, trucks may become the first area in which higher levels of autonomy become commonplace. Not in cities full of pedestrians and cyclists, but on long, repetitive logistics routes. There, technology, law and economics converge at a single point, creating conditions that passenger cars still do not have.
United States
It is currently the most advanced market for autonomous trucks.
Arizona, Texas, New Mexico
Since 2022–2024, autonomous trucks have been carrying out regular freight transport without a driver in the cab on selected highway routes.
• TuSimple conducted autonomous trips between logistics centers (e.g., Tucson–Phoenix). • Aurora Innovation is running commercial pilots for customers such as FedEx and Uber Freight. • Kodiak Robotics transports cargo for industrial and food companies along fixed routes.
In practice, this means that the truck drives autonomously on the highway, while oversight is provided from an operations center. Urban maneuvers are still limited or handled by a human.
China
China moved from testing to implementation very quickly.
Beijing, Shanghai, seaports Autonomous trucks operate: • in container ports, • in industrial zones, • on designated logistics routes.
Companies such as Baidu (Apollo) and Plus AI operate driverless transport without a safety driver, particularly in semi-closed environments where traffic is predictable.
Mines and Heavy Industry
This is the area in which autonomy functions for the longest time and most stably.
Australia, Canada, Chile • Autonomous mining trucks (e.g. from Caterpillar and Komatsu) have been operating for over 10 years. • They move without drivers, 24 hours a day. • They are centrally managed, with full control over routes and speed.
Under these conditions, autonomy proved safer and cheaper than human labor.
Europe
Europe is more cautious, but it is also implementing practical solutions.
Germany • Autonomous trucks tested and approved for use on fixed logistics routes. • 2021 – legal framework for Level 4 autonomy in freight transport.
Sweden • Autonomous electric trucks in ports and logistics centers (Volvo).
What benefits do autonomous cars offer users and cities
One of the main advantages of autonomous cars is the potential to improve safety. Human error, fatigue, inattention, and poor decisions are responsible for most road accidents. An autonomous system does not lose concentration, become tired, or give in to emotions. It analyzes situations continuously and consistently, significantly reducing the risk of collisions. Although the technology still has its limitations, the number of tests shows that in many cases, algorithms act faster than humans.
Another benefit is comfort. An autonomous car can become a space for work or relaxation, as the driver does not need to actively participate in driving. In cities, this may mean a new quality of mobility. Vehicles will move more smoothly, respond to one another, and minimize traffic jams. An intelligent autonomous fleet can also optimize routes, shortening travel times and reducing pollutant emissions.
It is also worth mentioning accessibility. Older people, people with disabilities, and those who cannot drive a traditional car will gain the opportunity to travel independently. This is not only a matter of convenience, but also of increasing independence and building more inclusive transportation.
What Challenges Face the Development of Autonomous Cars
Although technology is developing rapidly, the road to full autonomy is still a long one. One of the challenges involves legal issues. Current traffic regulations assume that the driver is responsible for the vehicle. In the case of an autonomous car, liability becomes less clear-cut. It must be determined whether responsibility for a potential accident lies with the user, the manufacturer, the software provider, or the communications network operator. These dilemmas require new regulations that balance safety, innovation, and citizens’ rights.
Practical Application of Ethics and Philosophy.
Another challenge is ethics. Autonomous systems must make difficult decisions in borderline situations. This raises the classic question of choosing between a lesser and a greater risk, between protecting the passenger and protecting pedestrians. An algorithm does not operate based on emotions, but on predefined rules and an analysis of consequences. That is why it is so important to establish clear principles for programming such decisions.
Technological challenges also remain significant. Despite advanced sensors and artificial intelligence models, situations still arise that are difficult for the system to interpret correctly. Problems may result from the unpredictable behavior of other road users, unclear road markings, or extreme weather conditions. Each additional scenario requires more data and model training, which is a time-consuming process.
From the perspective of city infrastructure, changes are also necessary to fully harness the potential of autonomy. This means investing in smart intersections, vehicle-to-infrastructure communication systems, and extensive data networks. Only by integrating onboard technology with urban infrastructure will autonomy become efficient and widespread.
History of autonomous cars
The history of autonomous cars is much longer than one might think and began well before the era of onboard computers and artificial intelligence. As early as the first half of the 20th century, engineers and visionaries wondered whether a vehicle could move without direct human control, although their ideas were based mainly on mechanics and infrastructure rather than software.
The first concepts emerged in the 1920s and 1930s. In the United States, experimental radio-controlled cars were presented that traveled along closed tracks. They were not autonomous in today’s sense, as they required a remote operator, but they demonstrated the very idea of separating the driver from the physical operation of the vehicle. In the 1950s and 1960s, visions of “the highways of the future” were popular, in which the car was to be guided by systems embedded in the road, such as wires or magnetic markers. The vehicle then became part of a larger, centrally controlled transportation system.
The real breakthrough came only with the development of electronics and computer science in the second half of the 20th century. In the 1980s, a team from the Bundeswehr University in Munich and the Ernst Dickmanns Institute built vehicles capable of independently staying in their lane and driving on motorways using cameras and computers. These were among the first examples of cars that “saw” the road and made decisions based on visual information rather than signals from infrastructure.
Another important stage was the period from 2000 to 2010, when government agencies and major technology companies became interested in autonomy. The DARPA Grand Challenges in the United States proved to be a turning point. The vehicles participating in them had to independently navigate long routes through desert terrain and later also in an urban environment. Although the first editions ended in spectacular failures, within just a few years the vehicles were able to complete entire routes without human intervention. It was then that the foundations of modern autonomous systems were established, including lidar, radar, sensor fusion, and route-planning algorithms.
In the years 2010–2020, development moved from laboratories to public roads. Companies such as Google (later Waymo), Tesla, Uber, and traditional car manufacturers began testing autonomous vehicles in real-world traffic. During this period, the first driver-assistance systems to become widely available in production cars also emerged, such as adaptive cruise control, automatic emergency braking, and lane-keeping assist. Although these were not fully autonomous systems, they represented an important step in getting drivers accustomed to handing over part of the control to machines.
Recent years have been a period of specialization and limited autonomy. Instead of attempting to create a car “for everything,” manufacturers have focused on specific scenarios, such as highway driving, navigating traffic jams, or autonomous taxis in selected cities. The systems have started to perform increasingly well, but at the same time it has become clear that full autonomy in all conditions is far more difficult than initially assumed.
Waymo, or What If Google Made Cars
Waymo is often perceived as an independent player in the self-driving car market, but its roots are closely tied to Google, and without understanding this connection, it is difficult to properly assess the scale and direction of this technology’s development. The self-driving project was launched within Google in 2009 as a research experiment aimed at determining whether advanced machine-learning algorithms could realistically handle road traffic in the real world.
In the early years, work on autonomous vehicles was carried out within the structures of Google X, the division responsible for the most ambitious and risky technology projects. It was there that the first prototypes of cars were developed, which drove hundreds of thousands of kilometers on public roads in California. The technologies used included those that Google had already been developing, such as highly accurate maps, image recognition systems, and the computing infrastructure needed to train artificial intelligence models.
In 2016, the project was spun off from Google and named Waymo, becoming a separate company. Formally, Waymo is no longer directly owned by Google, but by Alphabet Inc., the holding company that brings together various companies originating from Google. This separation was organizational and legal in nature, but it did not mean a severing of technological ties. Waymo continues to benefit from the Alphabet ecosystem, including the expertise of Google teams in artificial intelligence, data processing, and scaling real-time systems.
Waymo’s connection with Google was of enormous importance to the pace of its development. Access to global mapping data, data center computing power, and an engineering culture focused on experimentation enabled the company to move more quickly from prototypes to real-world services, such as autonomous rides through Waymo One. Waymo has become one of the best examples of how a research project within Google can evolve into an independent company that sets the direction for the development of the entire autonomous mobility industry.
How does autonomous vehicle law work?
The law governing autonomous cars did not emerge suddenly as a coherent set of regulations ready for a new technology. It was introduced gradually, often reactively, in response to the first tests, accidents, and pressure from manufacturers and technology companies. Legislators had to confront a problem they had not faced before: how to regulate a vehicle that is formally a car but, in practice, makes decisions autonomously.
At the beginning, that is, before 2010, traffic laws in most countries did not provide for the existence of a driverless vehicle at all. Regulations assumed that there was always a person behind the wheel who controlled the vehicle, was responsible for its maneuvers, and bore full responsibility for the consequences of driving. The first tests of autonomous cars therefore took place in a legal gray area, often on the basis of individual administrative permits or as part of closed research projects.
The breakthrough came with the development of road testing in the United States. Individual states, beginning with Nevada in 2011, started introducing the first regulations permitting the testing of autonomous vehicles on public roads. The law did not yet recognize autonomy as a fully legitimate means of operating a vehicle, but it allowed test driving provided that an operator was present, safety procedures were in place, and incidents were reported. This was an important signal that lawmakers were beginning to recognize a new category of technology.
In Europe, the process proceeded more slowly and cautiously. The Vienna Convention on Road Traffic was a significant obstacle for a long time, as it clearly stated that the driver must have full control of the vehicle. Only its interpretation and subsequent amendments made it possible to introduce systems that temporarily take over driving, provided that the driver can regain control at any time. This paved the way for the legalization of advanced driver assistance systems and partial autonomy.
The next stage involved the emergence of regulations concerning specific levels of autonomy. The law began to distinguish between situations in which a car merely assists the driver and those in which the system actually assumes responsibility for driving under specific conditions. Examples include regulations permitting autonomous driving on motorways or in traffic jams at limited speeds, where the risk of random events is lower and easier to predict.
At the same time, questions arose about legal liability. Who is responsible for an accident: the driver, the car manufacturer, the software provider, or perhaps the system operator? In many countries, the law still places responsibility on the human, even if the car was driving itself. At the same time, work is underway on product liability models that may, in the future, shift some of the risk onto manufacturers and algorithm developers.
The entry of law into the world of autonomous cars is therefore an evolutionary process, much like the technology itself. First, approval for testing, then the limited legalization of specific functions, and only later an attempt to create a coherent framework for vehicles that truly do not need a driver. This cautious pace shows that the law is not so much holding back autonomy as trying to keep up with technology that is changing the fundamental assumptions of road traffic.
Chronology
United States 2011 – Nevada becomes the first state in the world to legalize testing autonomous vehicles on public roads. 2012–2015 – additional states (California, Florida, Michigan) introduce their own testing regulations. 2020+ – selected states permit commercial rides without a safety driver (e.g., Arizona for Waymo).
Germany 2017 – legalization of automated driving systems (conditional autonomy), assuming that the driver can take control. 2021 – Germany becomes the first country in the world to introduce a legal framework for Level 4 autonomy in specific areas (e.g., shuttles, logistics).
Japan 2019 – amendment to road traffic law enabling the testing and limited use of autonomous vehicles. 2020 – approval of Level 3 autonomy in production cars. 2023 – first autonomous public transport services in cities.
United Kingdom
2015 – launch of government testing programs on public roads.
2018 – formal legal framework for testing autonomous vehicles without a conventional safety driver.
2022 – announcement of comprehensive legislation on autonomous vehicles (to be implemented in stages).
France
2019 – legalization of autonomous vehicle testing on public roads.
2021 – authorization of autonomous driving in specific scenarios (e.g. public transport, fixed routes).
China 2018 – first local regulations (Beijing, Shanghai) permitting autonomous vehicle testing. 2021 – nationwide legal framework for testing and pilot autonomous services. 2023 – autonomous taxis in selected cities without a safety driver.
Sweden
2017 – change in legislation enabling the testing of autonomous vehicles on public roads.
2018+ – pilot projects involving autonomous buses and urban vehicles.
The Netherlands 2015 – among the first regulations in Europe enabling tests of autonomous vehicles on public roads. 2019 – extension of the regulations to include tests without a driver in the vehicle (remote supervision).
Singapore 2017 – legal framework for testing autonomous vehicles in the city. 2020 – regular pilot programs for autonomous taxis and buses in designated districts.
Poland
2018 – amendment to the Road Traffic Act enabling tests of autonomous vehicles on public roads (with the consent of the relevant authorities).
No authorization for higher levels of autonomy in regular road traffic.
Saudi Arabia
2018 (June 24) – lifting of the ban on women driving cars. On this day, women in Saudi Arabia were legally allowed to obtain a driving licence and drive a car independently for the first time. The decision had enormous social and symbolic significance, as for decades the country had been the only one in the world with a formal ban on women driving cars.
2019 – first regulations enabling autonomous vehicle testing, mainly as part of pilot projects and special zones.
2021–2022 – launch of broader autonomous vehicle testing in the NEOM megaproject, planned as a futuristic city based on transport automation and reduced conventional car traffic.
2023+ – announcements of the use of autonomous buses, taxis, and urban transport systems in new districts and infrastructure projects.
In the case of Saudi Arabia, the juxtaposition of these two dates is particularly telling. The country, which allowed women to drive cars only very late, began investing in autonomous transport technologies almost simultaneously.
The Most Likely Development Scenarios.
The future of autonomous cars will depend on the pace of technological development, legal regulations, and social acceptance. There is increasing discussion of a model in which private cars become less popular and are replaced by autonomous on-demand transport services. Such a system would reduce the number of vehicles on the roads, improve their utilization, and lower maintenance costs.
It may also be possible to create special lanes for autonomous vehicles. This would allow vehicles to travel faster and without disruption, increasing road capacity. In the longer term, cities may completely transform traffic management. Intelligent management systems could enable traffic signals to be adjusted dynamically and even make it possible to abandon traditional intersections in areas where autonomous vehicles predominate.
One of the scenarios also involves integrating autonomous vehicles with public transport. Autonomous minibuses, taxis, and delivery vehicles could form a cohesive transportation network that meets residents’ needs quickly and efficiently. Artificial intelligence will control traffic, collect data, and optimize routes.
The impact of technology on the labor market cannot be overlooked either. On the one hand, autonomy may reduce the demand for professional drivers. On the other, it will create new roles in the operation, design, and monitoring of systems. Thus, not only the way we travel will change, but also the structure of the economy.
Summary
Autonomous cars are becoming one of the most fascinating directions in the development of modern mobility. They combine advanced sensors, communication systems, and artificial intelligence to bring a new level of safety, comfort, and efficiency to the roads. Despite the many challenges, the technology is moving toward solutions that may become the standard in global transportation in the future. Autonomy is not merely an improvement to the car. It is a fundamental change in how we move around the world and how we define the role of technology in everyday life.