How AVs Work

Autonomous vehicles (AVs) have a computer that can control the speed and direction of the vehicle, either to assist a human driver or to perform all driving tasks independently.

These vehicles are equipped with sensors that are used to take in different kinds of information about the surrounding world. This data goes to the computer, which then has to detect all surrounding objects, make a motion plan, and adjust the vehicle’s speed and direction accordingly.  
Below, we describe some of the details of these sensors and the operations of the AV computer.


Sensors

The three most common types of sensors in AVs today are cameras, lidar, and radar. They all have their advantages and disadvantages, and a mix of sensors is used to maximize the safety of the autonomous vehicle.

Cameras

Cameras are used to rapidly take images of the world. The computer uses these images to read road signs and traffic lights and to identify vehicles, pedestrians, and objects.

Pros:

  • Can see high detail and color
  • Sensors are cheaper than other types of sensors
  • Can work in light inclement weather (rain, snow, fog, etc.) as long as road markings are clearly visible.

Cons:

  • Data processing is expensive, slow, and challenging 
  • Can be blinded by sudden bright light or glare
  • Does not work in the dark
  • Does not work in heavy inclement weather (rain, snow, fog, etc.)

Lidar

Lidar uses a laser pulse to measure the distance to an object. The laser can be rotated to rapidly create a full 360-degree map of all the objects around the vehicle.

Pros:

  • Reliably detects objects around vehicle 
without any image processing
  • Data processing is very cheap and quick
  • Can be used in total darkness
  • Can work in light inclement weather (rain, snow, fog, etc.)

Cons:

  • Cannot read road signs or see color
  • Sensors are expensive (though the costs have 
decreased significantly in recent years)
  • Can only detect objects at shorter distances 
(200 meters or less)
  • Does not work in heavy inclement weather (rain, snow, fog, etc.)

Radar

Radar uses a pulse of radio waves to detect the speed and distance of objects.Radar can be used for many purposes, but is most useful for detecting moving objects in the path of the vehicle (e.g., another vehicle or pedestrian).

Pros:

  • Reliably detects moving objects in the path 
of the vehicle without any image processing
  • Works well for long-distance object detection 
(less well at very short distances)
  • Data processing is very cheap and quick
  • Can be used in total darkness
  • Sensors are moderately cheap
  • Can work in light and heavy inclement weather 
(rain, snow, fog, etc.).

Cons:

  • Cannot read road signs or see color
  • Struggles to distinguish stationary objects (e.g., stopped vehicles, road signs and clutter, etc.) from the background
  • Narrow field of view. Cannot easily make a 3D object map

See a comparison of how lidar and cameras see the world.

Computer Processing

Object Detection & Classification

As the computer drives a vehicle, the raw sensor data needs to be transformed into a consistent understanding of the world around it. This involves separating foreground objects from the background, and classifying the objects the computer detects in the scene. For example, the vehicle might see three pedestrians across the street standing still, one pedestrian on the side of the road walking at a slow speed, and an oncoming truck traveling in the adjacent lane at 25 mph. All of these objects must be correctly detected and classified for the vehicle to operate safely.


A comprehensive mix of sensors gives the computer the best ability to classify objects. Things can go wrong when the computer only has camera data to work with. For example, in the image below, a computer tried to convert the camera data into a collection of objects. It mistakenly saw a large person standing on the road and did not correctly identify the large truck completely blocking the vehicle’s path. With lidar and radar data also available, the computer would be able to “see” the size and shape of the truck and would be much more likely to identify the whole truck as a truck. Cases like this are a major reason why autonomous vehicles should be guided by a diverse mix of sensors. ​

Image credit: Simon Stent, Toyota Research Institute, 2016

Computer Training: AI Learning and Mapping

AV computers are trained to drive by studying large data sets of previously recorded human driving. Human-driven test vehicles collect data by driving around to build up a repository of what different vehicles look like, what different pedestrians look like, etc. This data is labeled by a human (“vehicle,” “pedestrian,” “cyclist,” “stop sign,” etc. are common labels) and then fed into AI algorithms that learn to recognize these objects when they appear on the road. As more data is used for training and the algorithm improves, the AV learns to recognize all the different shapes, sizes, and colors of different vehicles, pedestrians, and other objects.

AVs are also guided by maps of the environments that they drive in. These maps include landmarks used to determine the vehicle’s position when the sensors see them again, as well as how the lanes in the road are connected. This helps the AV know where it’s safe to perform a lane change, which traffic lights control which lanes, and how to get from one street to another.

Most AV companies extensively map the areas they will operate in so the vehicle recognizes where it can and cannot drive. For instance, Chevy Supercruise can only be activated on highways that have been pre-mapped, and Waymo taxis only operate on city streets that have been pre-mapped. Other systems, such as Tesla Autopilot and Full Self-Driving, can be operated in areas that have not been pre-mapped, though this can be problematic if the vehicle makes poor decisions based on incorrectly interpreted sensor data.

Stay Updated

Sign up for our newsletter to get updates delivered 
straight to your inbox.

Goal 4

Close the Level 2+
Loophole

According to the Society of Automotive Engineers J3016 Standard, the human driver is considered to be in control of a vehicle using a driver assistance system (Levels 1 and 2 ADAS systems), while the computer driver is responsible for fully autonomous systems (Levels 3, 4 and 5 ADS systems).  

Unfortunately, a loophole in J3016 allows car companies to blur this distinction with so-called Level 2+ systems. These systems allow the computer driver to perform complex driving tasks, even completing entire trips without any human intervention, but in the event of a collision the vehicle manufacturer can pin the liability on the human driver.

Campaign Goal 4 is to close this Level 2+ loophole by addressing several issues:

  • Redefine Level 2+ as an ADS System: Legally define any automated driving system that is capable of departing a marked lane of travel as an ADS system. Require it to go through AV testing permitting process. (Similar legislation is currently being considered in California [CA SB 511]).
  • Level 2+ Liability: Assign criminal fines and civil liability to manufacturers of Level 2+ vehicles for any traffic law violations.
  • Truth In Advertising: Ban advertising and brand names of ADAS systems that falsely imply that the vehicles are fully autonomous.

Goal 3

Establish Manufacturer Crash Liability and Mandatory Crash Reporting​

Advanced driver assistance systems (Level 2 ADAS systems) currently operate with essentially no state-level regulations.

Unfortunately, civil and criminal liability law has not kept up with the proliferation of ADAS systems, and human drivers (not vehicle manufacturers) are almost always held liable for crashes involving the use of ADAS systems. Further, states do not currently require vehicle manufacturers to report ADAS crashes or crash data to state agencies. Companies are only required to report ADAS crashes to federal agencies, but this requirement was significantly reduced in 2025.

Campaign Goal 3 is to establish clear legal liability laws that protect the rights of human drivers while using ADAS systems and to require all companies to report all crashes and crash data involving ADAS systems to the state.

Specific legislative goals include:

  • Driver Monitoring: Assigns criminal fines and civil liability to manufacturers of Level 2 vehicles for any traffic law violations if the human driver was not engaged.
  • Operational Design Domain (ODD): Subjects manufacturers of Level 2 vehicles to civil penalties and a private right of action if their Level 2 system is able to operate outside of the ODD and if the manufacturer has not defined the ODD as the conditions in which the vehicle can operate safely.
  • Collision Reporting: Require vehicle manufacturers to report any ADAS collision to the state DMV within five days (similar legislation is currently being considered in California [CA SB 572]).
  • Data Deletion: Set a civil penalty of $10 million for vehicle manufacturers that destroy or hide “black box” data relevant to a crash investigation.

Goal 2

Taxi & Sensor Safety Regulations

Most autonomous vehicles in development today are guided by a diverse mix of sensors, including cameras, radar, and lidar (see “How AVs Work” for more details). Unfortunately, some companies are pursuing single-sensor (i.e., camera-only) designs to minimize vehicle costs.

Taxis have to operate in complex urban environments, with frequent road surprises like vehicles, cyclists, and pedestrians unexpectedly entering the vehicle path. Driver assistance systems cause drivers to pay less attention to the road, increasing the risk of collisions. Self-driving cars require a mix of sensors to maximize their ability to detect and avoid road hazards.   

Campaign Goal 2 is to make sure all autonomous vehicles use a mix of sensor technologies to ensure maximum vehicle safety.

The following regulations can be implemented via local taxi regulators and municipal legislation in addition to state legislation.

  • Taxi Licensing – Single-Sensor Regulations: Prohibit vehicles used as licensed taxis from activating single-sensor ADAS and ADS systems. Require a mix of sensors, including lidar.
  • Geofencing: Place restrictions on the operation of single-sensor ADAS and ADS taxis, including the operation of vehicles near school zones and work zones, in adverse driving conditions, and outside that vehicle’s pre-established operational design domain.
  • Municipally Owned Vehicles: Municipalities can commit to not purchasing vehicles equipped with single-sensor ADAS and ADS systems.

Video: An ADAS-equipped vehicle with no lidar sensor fails to avoid an overturned truck. No one was hurt in this collision.

Goal 1

Establish California-Style AV Regulations in 
More States

California leads the nation in both the development and regulation of fully autonomous vehicles (Levels 3, 4 and 5 ADS systems). California-based AV companies like Waymo, Nuro, and Zoox have developed their technology while complying with the rigorous regulations imposed by the state.

Around 30 states allow the testing and/or deployment of fully autonomous vehicles, with a wide range of data reporting requirements, testing permit processes, and liability definitions. Major states such as Illinois and New York do not presently have state laws authorizing the testing of ADS systems.

Campaign Goal 1 is to pass California-style AV regulations in many additional states to establish rigorous safety standards across the nation.

Important features of the California regulations include:

  • Three stages of development permits: testing with a driver in the vehicle; testing without a driver in the vehicle; and full deployment.
  • Required definition of the operational design domain (the geographic areas and weather conditions in which the vehicle will be operated).
  • Regular crash and safety data reporting to the state DMV.
  • A ban on misleading advertising of the safety and capabilities of AVs.

 

These regulations ensure that vehicles are thoroughly tested and shown to be safe before they can be used in a robotaxi service or as a personal autonomous vehicle.

These regulations are available online through the California DMV.