ADAS (advanced driver-assistance systems)

Advanced driver-assistance systems, universally shortened to ADAS, are the suite of technologies that help a human drive more safely without taking over the driving entirely. They sense the road and surrounding traffic and then either warn the driver or briefly take partial control to prevent or mitigate a problem. Familiar examples include automatic emergency braking, which brakes when a collision is imminent, lane-keeping assistance, which nudges the car back into its lane, adaptive cruise control, which maintains a safe following distance, and blind-spot warnings. The common thread is assistance: the driver stays in charge and remains responsible, while the system acts as a vigilant helper.

This is what places ADAS below full self-driving on the spectrum of vehicle automation. Rather than replacing the driver, ADAS supports a driver who must remain engaged, which distinguishes it from a fully automated driving system that performs the entire driving task. To do its job, ADAS depends on perception built from sensors such as cameras and radar, and sometimes LiDAR, fused together to detect lanes, vehicles, pedestrians, and hazards. In practice, ADAS is where a great deal of real-world visual and sensor AI already operates at scale, quietly assisting drivers in ordinary cars.

ADAS builds a picture of the driving environment from sensors, typically fusing cameras and radar, and sometimes LiDAR, to detect and track lanes, vehicles, pedestrians, and obstacles. Software interprets this picture to recognize situations that call for a response, such as a vehicle braking ahead or the car drifting out of its lane. Depending on the feature, the system then either alerts the driver or applies a limited, specific control action like braking or steering assistance. Crucially, these actions are bounded and supervisory, designed to support a driver who is expected to stay attentive and take over as needed.

ADAS matters because it is one of the most widespread real-world deployments of visual and sensor AI, actively improving road safety in millions of ordinary vehicles today. For anyone interested in physical AI, it represents the practical, already-shipping end of vehicle automation, distinct from the more aspirational goal of full self-driving. Understanding ADAS, and how it assists rather than replaces the driver, clarifies where much of automotive perception currently lives and why the boundary between assistance and autonomy is such an important distinction.

ADAS assists a human driver who remains responsible for the vehicle, providing warnings or limited control for specific tasks. A fully self-driving system performs the entire driving task itself within its operating conditions. ADAS supports the driver, whereas full autonomy replaces the driver.

Common features include automatic emergency braking, lane-keeping and lane-departure assistance, adaptive cruise control, blind-spot monitoring, and parking assistance. Each targets a specific safety or convenience task while leaving overall control with the driver.

ADAS typically fuses cameras and radar, and in some vehicles LiDAR, to perceive lanes, vehicles, pedestrians, and hazards. Combining sensors lets the system remain reliable across a wider range of conditions than any single sensor would allow.

Radar, LiDAR, Sensor fusion, Ego vehicle, Operational design domain (ODD).