Radar-Centric ADAS

A radar-centric approach to ADAS has demonstrated advantages over a camera-centric approach

Ready for All Kinds of Conditions


As Aptiv’s environmental chamber at CES demonstrates, using radar with machine learning to build a sensing and perception platform makes a lot of sense. A radar-centric system works well in a variety of weather and lighting conditions – conditions where cameras and human vision encounter difficulty – while operating 25 percent more cost-effectively and 65 percent more energy-efficiently than camera-centric systems.

Superior in Various Scenarios

Here’s just a sampling of the use cases where a radar-centric approach shines.
Vulnerable Road Users at Night

Because it doesn’t rely on light, radar is better at detecting vulnerable road users at night, and machine learning helps radar-centric systems predict their behavior and avoid false positives and negatives.
Watch video

Stop-and-Go Traffic

Machine learning enables a radar-centric system to accurately track stationary vehicles, while Aptiv’s legacy of automated driving gives us the experience and motion-planning tools to fine-tune vehicle control at low speeds.
Watch video

Automatic Lane Change

Aptiv’s corner radars provide a robust, 360-degree view – including the ability to “see through” obstructions in some cases – which allows the vehicle to accurately track target vehicles.
Watch video

Low-Speed Cut-Ins and Cut-Outs

The system must maintain a safe distance from other vehicles without “overreacting” when they move into or out of the lane. Aptiv’s system can quickly detect and then consistently and accurately track them, while our vehicle-control algorithms respond in a natural but assertive way.
Watch video

Stationary and Clustered Objects

While stationary and tightly clustered objects would traditionally pose a challenge to radar-based ADAS, machine learning allows the system to accurately track those objects, as well as discriminate clustered pedestrians.
Watch video

Level 4 Valet Parking

Radar-centric ADAS can clearly see all the stationary vehicles and vulnerable road users in a parking lot or busy street and detect empty parking spots much farther away than vision-only systems, allowing the vehicle to start the parking process before it reaches the space.
Watch video

Lane Centering Without Markers

When roads don’t have visible lane markings, Aptiv’s system uses Radar Based Localization to maintain appropriate spacing and biasing.
Watch video


AI/ML Gets the Most from Radar

A radar-centric approach to ADAS offers the broadest possible operating performance and provides a compute- and power-efficient solution that democratizes advanced safety features by lowering system costs.

Learn more

Automotive Radar Technologies Sharpen Focus

Recent advances in vehicle radar technology are about to create a step change in capabilities that will greatly strengthen a radar-centric approach to ADAS.

Learn more
Automotive Radar Technologie
Machine Learning Improves Perception

How Machine Learning Improves Perception

Machine learning is a subset of artificial intelligence that refers to a system’s ability to be trained through experience with different scenarios. One challenge machine learning helps address with radar is edge detection.


Learn more