Technological progress is rapidly changing the face of modern transport, and the question of what driverless cars are called is no longer the stuff of science fiction. Today on the roads you can find prototypes that are controlled by complex algorithms, but in everyday life many different definitions are used for them. Most often, such vehicles are called driverless cars, however, the term covers a wide range of technologies.
Depending on the degree of human participation in the management process, experts distinguish different classifications. Some systems only help the driver park or keep a lane, while others are able to take full control of the situation on the highway. This is why it is important to distinguish between the concepts autopilot and full autonomous driving.
In this article we will look in detail at what terminology exists, how the levels of autonomy differ according to the SAE standard, and what technologies are hidden behind the beautiful names. Understanding these differences will help you better navigate auto industry news and correctly assess the capabilities of modern technology.
Basic Terminology: Drone or Robot?
The most common and technically accurate term is driverless car. This is a general name for any vehicle capable of navigating its environment and moving without constant human control. In a professional environment, the abbreviation BTS is often used - unmanned vehicle.
However, in media and marketing you can find other names that sometimes mislead users. For example, the term “robomobile” is more often associated with specific projects like Google Waymo, where the car was originally designed without a steering wheel and pedals. It is important not to confuse marketing names with technical essence.
- 🚗 Self-driving car - a general technical name for any car with autonomous control systems.
- 🤖 Robomobile - often used to refer to fully autonomous machines without manual controls.
- 🧠 Autonomous vehicle (Autonomous Vehicle) is a term that emphasizes the ability of AI to make decisions.
It is worth noting that the concept full autonomy implies the absence of the need for human intervention even in emergency situations. At the same time, many modern systems, which manufacturers call “autopilots,” are not such and require constant driver attention.
The diversity of terms arises because technology develops faster than language can standardize. Lawyers, engineers, and marketers use different words to describe the same functionality, which creates some confusion for the end user.
SAE autonomy levels
To systematize knowledge about how driverless cars are called depending on their capabilities, engineers developed the international classification SAE J3016. It divides all systems into six levels, from zero to five. Understanding these levels is critical to safety.
At the initial stages, from level 0 to level 2, the person remains primarily responsible for management. The systems only assist: cruise-control, lane keeping or automatic parking. It’s too early to use the term “car without a driver” here, since control is entirely human.
⚠️ Attention: The use of Level 2 systems (for example, Tesla Autopilot) does not relieve the driver of responsibility. You must keep your hands on the steering wheel and watch the road, even if the car is driving itself.
Starting from level 3, the situation changes. The car can take over control in certain conditions (for example, on the highway in traffic), but requires human intervention upon request. This is a transitional stage where terminology becomes especially important: this is no longer just an assistant, but not yet a full-fledged robot.
Levels 4 and 5 are what we used to call full-fledged drones. At level 4, the car can drive itself in a limited geo-zone or under certain weather conditions without human intervention. Level 5 is the ideal when the car goes everywhere and always without any intervention.
| Level | Title | Who controls? | Do you need a driver? |
|---|---|---|---|
| 0 | No automation | Man | Yes, all the time |
| 1 | Assistant | Man + System | Yes, all the time |
| 2 | Partial automation | System (under control) | Yes, all the time |
| 3 | Conditional Automation | System (in some cases) | Yes, but not always |
| 4 | High automation | System | No (in range) |
| 5 | Full automation | System | No, never |
It is at levels 4 and 5 that a car can rightfully be called robotaxi or an autonomous shuttle. In these cases, the concept of the “driver” as such disappears, giving way to the passenger. Engineers continue to work on the transition from Level 3 to Level 4, as this is the most difficult technological barrier.
What is the main difficulty in moving from level 3 to level 4?
The transition from Level 3 to Level 4 is challenging because the system must be able to safely exit or park if the driver is unresponsive or if environmental conditions are beyond the capabilities of the sensors. This requires redundancy in all systems.
Technologies that provide autonomy
Behind the car’s ability to move without a driver is a complex set of hardware and software solutions. The basis is a sensory unit that replaces human eyes and ears. Lidars, radars and cameras create a digital map of the surrounding space in real time.
Lidars (Light Detection and Ranging) emit laser pulses and measure their return time, creating an accurate 3D model of the area. Radars work great in bad weather, determining the speed of objects, and cameras read road markings and signs. All this data is processed by powerful on-board computers.
- 📡 Lidars — create a three-dimensional map of the environment with high accuracy of distances.
- 📹 Computer vision — neural networks analyze the video stream, recognizing pedestrians and traffic lights.
- 🛰️ GLONASS/GPS — provide precise positioning of the vehicle on the map with centimeter accuracy.
The software part is no less important. Artificial intelligence must not only see objects, but also predict their behavior. For example, the algorithm must understand that a child at the edge of the road can suddenly run out, even if he is currently standing still. This requires training over millions of kilometers.
Communication between components is also critical. Modern systems use high-speed data buses to transfer information from sensors to the processor. Any delay in data transmission can cost lives, so the reliability of the architecture on-board computer is priority number one.
When choosing a car with autopilot, pay attention not only to the name of the function, but also to the type of sensors used. The presence of lidar often indicates a more advanced level of the system.
Legal status and responsibility
The question of what driverless cars are called is closely related to legal aspects. In most countries, legislation has not yet fully adapted to the emergence of levels 4 and 5 of autonomy. Who is responsible in the event of an accident: the owner, the manufacturer or the software developer?
Currently, if a car has a steering wheel and pedals, the person sitting behind the wheel is legally responsible for the controls. Even if the autopilot is turned on, the driver is required to control the situation. This creates a legal conflict: the system controls, but responsibility lies with the individual.
⚠️ Attention: In the event of an accident involving the autopilot system, insurance companies will first request system logs. Do not try to modify the car software yourself - this may be grounds for refusal of payment.
For fully autonomous taxis, where the driver is physically absent, responsibility gradually shifts to the service operator or technology manufacturer. Some US states and EU countries have already passed laws allowing the operation of such vehicles without a person in the cabin, but only in test areas.
The development of telematics makes it possible to transmit data about the state of the vehicle and system actions to the cloud in real time. This helps resolve controversial situations. Lawyers are developing new concepts of “electronic identity” for AI, but the implementation of such standards is still far away.
Safety and ethical dilemmas
The introduction of driverless cars raises not only technical, but also ethical questions. The famous “trolley problem” confronts algorithm developers: who should be saved in a hopeless situation? The software code must be programmed to make moral decisions in a split second.
From a statistical point of view, artificial intelligence does not get tired, does not get drunk and is not distracted by the phone, which should theoretically reduce the number of accidents. However, cybersecurity is becoming a new battle front. Hacking a drone's control system can lead to catastrophic consequences.
- 🛡️ Cyber protection — protection of communication channels from hacker attacks and control theft.
- ⚖️ AI ethics — decision-making algorithms in emergency situations.
- 🚧 Testing — the need for billions of kilometers of virtual and real tests.
Manufacturers conduct thousands of hours of simulation to train a car to respond to rare events. But no simulation can capture all the chaotic situations in the real world. That is why the transition period, when there are people and robots on the roads at the same time, is the most dangerous.
It is important to understand that we have a long way to go before reaching level 5 autonomy, when the car can ignore human errors. Public trust in the technology is growing slowly, and every incident involving a drone sets the industry back.
☑️ What to check before using autopilot
The future of transport: from ownership to service
The advent of full-fledged driverless cars is changing the very concept of car ownership. Why buy an expensive device that sits idle 95% of the time if you can call a robotaxi with a button? Model MaaS (Mobility as a Service) is becoming dominant in urban studies.
Cities of the future are planned taking into account the fact that parking will no longer be necessary, since cars will be constantly in motion, carrying passengers. This will free up huge spaces for parks and pedestrian areas. Traffic will become smoother by connecting vehicles into a single network (V2X).
However, mass adoption will face social challenges. Millions of taxi, truck and bus drivers could lose their jobs. Society will have to find a way to retrain these people or introduce the concept of unconditional income. The technological gap will be acutely felt.
The main trend of the future is the transition from personal car ownership to the use of transport as a service (Taxi-hailing), which will change the appearance of cities and the economics of transport.
In conclusion, driverless cars are not just a gadget, but a fundamental shift in civilization. Regardless of what we call them - robo-cars, autonomous cars or drones - they are already here. And our task is to learn to coexist safely with them.
Frequently asked questions (FAQ)
Is it possible to sleep at the wheel if autopilot is on?
No, absolutely not. Modern systems (levels 1-3) require constant control by the driver. Cameras in the cabin track the direction of your gaze, and sensors on the steering wheel track the presence of hands. Sleeping behind the wheel, even with Autopilot engaged, can lead to an accident and legal liability.
How does a car without a driver understand the traffic controller's signals?
This is one of the most difficult tasks. Cameras read gestures, and computer vision algorithms interpret them. However, in complex situations where signals are inconsistent, the drone may stop and request assistance from a remote operator or wait for clearer ones.
Do drones work in winter when it snows?
In its current form - with restrictions. Snow and ice can gum up camera lenses and distort lidar readings. Many Level 4 systems have weather restrictions and may refuse to operate or ask the driver to take over control if visibility becomes poor.
Who is to blame if a self-driving car hits a pedestrian?
Responsibility depends on the level of autonomy and circumstances. If it was level 2 or 3 and the driver ignored the warnings, the driver is at fault. If it was a completely offline mode (level 4-5) and a failure occurred in the software, the manufacturer or service operator is responsible.
When will driverless taxis appear in every city?
Forecasts vary. Limited operation in large metropolitan areas is already underway. Mass appearance in every city is expected no earlier than 2030-2035, when the technology becomes cheaper and legislation is fully adapted.