Mobility technology company AVL has released the Ground Truth Reference System to help enable the efficient and reliable development and validation of vehicle perception sensor technology. The system utilizes the Deepen Calibrate sensor calibration suite from startup Deepen AI, which enables precise sensor calibration during driving operations and the accurate collection of ground truth data.
Depending on the vehicle’s degree of automation – from SAE Level 0 to 5 – more software-controlled sensor systems will be required to deal with environment detection and how to respond to certain elements. A multitude of sensors is used to build the perception layer of advanced driver assistance systems (ADAS) for autonomous driving, and these include lidar, cameras and radars to provide the vehicle with an accurate view of its surroundings during different scenarios.
AVL Ground Truth Reference System, otherwise known as DGT, is mounted on a vehicle to capture a holistic 360° field of view of the vehicle’s static and dynamic surroundings. DGT recorded ground truth data is gathered from lidar, camera and GPS sensors to ensure an accurate view of the vehicle and its systems’ test environments. Engineers are then able to conduct a statistical analysis of large amounts of representative, accurate real-world driving data to assess the performance of the ADAS/AD sensors and the perception algorithms.
“Independent, high-precision ground truth reference systems such as AVL DGT play an increasingly critical role for our OEMs to objectively develop and validate their vehicles sensor performance,” said Thomas Guntschnig, portfolio manager ADAS/AD testing solutions, AVL. “With Deepen Calibrate, we can ensure that our AVL DGT-sensors are calibrated with high accuracy in a fraction of time during driving campaigns.
“Reliable data is the foundation for safety. By combining our efforts at Deepen AI through Deepen Calibrate and AVL’s cutting-edge AVL Dynamic Ground Truth solution, we can significantly increase the quality of data and increase the safety of ADAS/AD systems,” said Mohammad Musa, CEO and co-founder, Deepen AI.