Distortion in temporal sensor tracks is used to infer object velocity and bounding boxes, improving autonomous vehicle perception accuracy.
Adjacent storage diodes and alternating modulation gates speed charge transfer, cut temporal noise, and improve TOF depth precision.
Shared processing of multiple polarization return signals helps LIDAR identify materials while reducing duplicated electronics and ADC count.
Rear-facing FMCW LIDAR on the tractor tracks trailer yaw, roll, pitch, and position to trigger corrective action before instability grows.
A polar occupancy grid and probabilistic motion model improve pedestrian path prediction accuracy while preserving interpretability for AV collision avoidance.
A SPAD correlation sensor measures LED-coded light delay to produce distance images with lower computation, power use, and integration cost.
Adaptive LiDAR thresholding uses repeated comparator outputs to reject solar noise while preserving detection of distant or low-reflectivity objects.
A split coherent optical path enlarges LiDAR reception aperture and reduces crosstalk, improving echo energy for more accurate speed and distance sensing.
Two LiDAR sensors compare clustered range data to suppress fog, snow, exhaust, and debris returns that can mislead vehicle navigation.
Modality-based sensor plugins normalize and validate radar, lidar, and camera data before grid map fusion for more reliable autonomous control.
A prismatic-metalized reflective surface boosts radar, lidar, and camera returns from cyclists and pedestrians to improve collision avoidance.
Polarization-diverse reflections are handled through shared transform processing, reducing duplicate electronics while preserving LIDAR material identification.
A split secondary beam is directed toward the receive path to widen near-field coverage, reduce parallax error, and improve object detection.
Filters, adaptive emission control, and shuttering protect LIDAR from external light that can cause detector damage and false scan data.
An asymmetrical loss model separates out-of-plane radar returns from in-plane Doppler data to improve vehicle velocity estimation.
Vehicle strobe pulses and passive retroreflectors replace heavier radar hardware to guide UAM aircraft to specific landing pads.
A reflective periscope extends LIDAR coverage into ultra-nearfield blind spots, improving object detection around autonomous vehicles.
Frequency-modulated lidar with coherent detection improves range and velocity sensing in bright sunlight while reducing crosstalk and self-interference.
A polarization grating switches one laser beam between multiple scanners for precise routing, faster scanning, and stronger LiDAR signal quality.
Processors write lidar photodetector data into predetermined stream locations, avoiding shared buffers to cut latency and storage.