How are autonomous vehicles tested?

Testing AVs is a complex and evolving challenge. AVs are designed to navigate their environment independently, using advanced sensor technologies such as LiDAR, radar, cameras, and ultrasonic sensors. These sensors provide the AVs with a 360-degree perception of their surroundings, enabling them to handle complex scenarios without human intervention. To ensure the safety and performance of AVs, they need to be tested under various conditions and situations, ranging from simple to increasingly complex. AB Dynamics’ track testing products allow even the most complex scenarios to be created and repeated safely on the track.

Repeatable and accurate control

Our extensive range of driving robots are instrumental in the development and testing of AVs. They are used to develop the control algorithms used in self-driving vehicles. In what we term “external follow mode”, these robots act as actuators, responding to real-time control signals. This allows engineers to test and optimise self-driving software using any car, making driving robots a versatile and effective solution for AV testing.

ADAS targets and platforms

In AV testing, unexpected traffic situations such as a driver running a red light or an emergency vehicle approaching rapidly present unique challenges. Our solution involves using realistic ADAS targets and robust powered platforms to simulate real-life scenarios and to test edge cases. Our extensive range of targets and target-carrying platforms can mimic vehicles, pedestrians, and motorcyclists. Platforms can be synchronised not only with each other but also with vehicles under test, creating a dynamic and realistic testing environment. ADAS platforms utilise the same powerful RC software as our driving robots, simplifying synchronisation and path creation. Managing paths and speed profiles is straightforward and they can be easily transferred between driving robots, targets and platforms.