Autonomous Vehicles – Farzad Cyrus Foroughi Abari, Forrest Samuel Briggs, Alexander Thomas Starns, Lyft Inc
Abstract for “Sensor calibration laboratory”
“In one embodiment, the facility for calibrating sensors for an autonomous vehicle (AV), includes a camera calibration target that can be used to calibrate an optical camera; a light detection & ranging (LiDAR); calibration target that can be used to calibrate a LiDAR transceiver; and a platform that allows the AV to park and drive on the platform. While the vehicle is parked on the platform, the camera calibration target and LiDAR calibration targets are placed so that they can be detected by the optical camera as well as the LiDAR transceiver. While the AV is parked on a platform, the platform can modify the lateral position, height or orientation of the LiDAR transceiver and optical camera relative to the LiDAR calibration targets and the camera calibration target.
Background for “Sensor calibration laboratory”
A safe and successful navigation system for an AV requires accurate measurement and representation of its external environment. A three-dimensional model or data of the external environment that is accurate and reliable depends on the accuracy of the AV sensors. This may include understanding the initial configurations of the sensors as well as how obstacles and objects are detected by them. The sensors of an AV might need periodic calibration to maintain an accurate representation of the environment. This is necessary for safe navigation. Recalibration can be time-consuming due to the many sensors an AV might be equipped with. While the sensors are being calibrated, AVs can be inoperable or taken offline. This reduces the use of the AVs as well as the capabilities of an AV management software. Calibration of sensors can be a tedious, manual process, which may require specialized labor, and increase costs for operation.
Summary for “Sensor calibration laboratory”
A safe and successful navigation system for an AV requires accurate measurement and representation of its external environment. A three-dimensional model or data of the external environment that is accurate and reliable depends on the accuracy of the AV sensors. This may include understanding the initial configurations of the sensors as well as how obstacles and objects are detected by them. The sensors of an AV might need periodic calibration to maintain an accurate representation of the environment. This is necessary for safe navigation. Recalibration can be time-consuming due to the many sensors an AV might be equipped with. While the sensors are being calibrated, AVs can be inoperable or taken offline. This reduces the use of the AVs as well as the capabilities of an AV management software. Calibration of sensors can be a tedious, manual process, which may require specialized labor, and increase costs for operation.
Click here to view the patent on Google Patents.