Pony Ai Inc.

Pony Al Inc.

Yubo Zhang

Xiang Yu

Jun Peng

Tiancheng Lou

Nengxiu Deng

Kai Chen

Cheng Jin

Zuoteng Chen

Zhongnan Hu

Pingfan Meng

Hao Song

Sinan Xiao

Tianyi Li

Tao Qin

Yiming Liu

Yin Zhong

Stephen Lee

Wei-Yang Chiu

Described herein are systems, methods, and non-transitory computer readable media for memory address encoding of multi-dimensional data in a manner that optimizes the storage and access of such data in linear data storage. The multi-dimensional data may be spatial-temporal data that includes two or more spatial dimensions and a time dimension. An improved memory architecture is provided that includes an address encoder that takes a multi-dimensional coordinate as input and produces a linear physical memory address. The address encoder encodes the multi-dimensional data such that two multi-dimensional coordinates close to one another in multi-dimensional space are likely to be stored in close proximity to one another in linear data storage. In this manner, the number of main memory accesses, and thus, overall memory access latency is reduced, particularly in connection with real-world applications in which the respective probabilities of moving along any given dimension are very close.

Memory architecture for efficient spatial-temporal data storage and access

Described herein are systems, methods, and non-transitory computer readable media for constructing and utilizing vehicle field-of-view (FOV) information. The FOV information can be utilized in connection with vehicle localization such as localization of an autonomous vehicle (AV), sensor data fusion, or the like. A customized computing machine can be provided that is configured to construct and utilize the FOV information. The customized computing machine can utilize the FOV information, and more specifically, FOV semantics data included therein to manage various data and execution patterns relating to processing performed in connection with operation of an AV such as, for example, data prefetch operations, reordering of sensor data input streams, and allocation of data processing among multiple processing cores.

Systems and methods for constructing and utilizing field-of-view (FOV) information

A computer-implemented method and a system for training a computer-based autonomous driving model used for an autonomous driving operation by an autonomous vehicle are described. The method includes: creating time-dependent three-dimensional (3D) traffic environment data using at least one of real traffic element data and simulated traffic element data; creating simulated time-dependent 3D traffic environmental data by applying a time-dependent 3D generic adversarial network (GAN) model to the created time-dependent 3D traffic environment data; and training a computer-based autonomous driving model using the simulated time-dependent 3D traffic environmental data.

Generative adversarial network enriched driving simulation

A sensor enclosure comprises a cover and a structure. The structure can be encased by the cover. The structure comprises a frame, a ring, and one or more anchoring posts. The frame can be configured to mount one or more sensors. The ring, disposed peripherally to the frame, can be operatively coupled to the cover. The ring can include a drainage ring plate that drains rainwater accumulated on the cover away from the sensor enclosure. The one or more anchoring posts, disposed underneath the frame and the ring, can be used to anchor the sensor enclosure to a vehicle.

Sensor enclosure drainage

Described herein is a memory architecture that is configured to dynamically determine an address encoding to use to encode multi-dimensional data such as multi-coordinate data in a manner that provides a coordinate bias corresponding to a current memory access pattern. The address encoding may be dynamically generated in response to receiving a memory access request or may be selected from a set of preconfigured address encodings. The dynamically generated or selected address encoding may apply an interleaving technique to bit representations of coordinate values to obtain an encoded memory address. The interleaving technique may interleave a greater number of bits from the bit representation corresponding to the coordinate direction in which a coordinate bias is desired than from bit representations corresponding to other coordinate directions.

Dynamic memory address encoding

Described herein are systems, methods, and non-transitory computer readable media for performing an alignment between a first vehicle sensor and a second vehicle sensor. Two-dimensional (2D) data indicative of a scene within an environment being traversed by a vehicle is captured by the first vehicle sensor such as a camera or a collection of multiple cameras within a sensor assembly. A three-dimensional (3D) representation of the scene is constructed using the 2D data. 3D point cloud data also indicative of the scene is captured by the second vehicle sensor, which may be a LiDAR. A 3D point cloud representation of the scene is constructed based on the 3D point cloud data. A rigid transformation is determined between the 3D representation of the scene and the 3D point cloud representation of the scene and the alignment between the sensors is performed based at least in part on the determined rigid transformation.

Sensor alignment

A sensor enclosure comprising a domed cover and a base. The base can be encased by the domed cover. The base comprises an inner frame, an outer frame, one or more wipers, and a powertrain. The inner frame can provide surfaces for one or more sensors. The outer frame, disposed underneath the inner frame, the outer frame includes a slewing ring. The slewing ring comprises an inner ring to which the domed cover is attached and an outer ring attached to the outer frame. The one or more wipers extends vertically from the outer frame, each wiper having a first end attached to the outer frame and a second end attached to a support ring, and each wiper making a contact with the dome cover. The powertrain, disposed within the outer frame, configured to rotate the ring and the dome cover attached to the inner ring.

Alternative wiper mechanism to remove rainwater on a sensor enclosure

Systems, methods, and non-transitory computer readable media may be configured to characterize optical characteristics of optical elements. An optical element mount may be configured to carry an optical element. A calibration display may be configured to display a calibration object. The calibration object may include a known visual pattern. Multiple images of the calibration object may be obtained. The multiple images may be acquired using the optical element carried by the optical element mount. The multiple images may include different perspectives of the calibration object. Optical characteristics of the optical element may be characterized based on the known visual pattern and the different perspectives of the calibration object.

Characterizing optical characteristics of optical elements

A computer implemented method for operating a wiper system of a sensor enclosure. The wiper system can be configured to receive a camera operating information from one or more cameras. The one or more cameras can be determined to be in an exposure mode based on the camera operating information. A first speed of one or more wipers can be adjusted such that the one or more wipers are not in field of views of the one or more cameras while the one or more cameras are in the exposure mode.

Method for controlling wipers based on a camera exposure time

A computer implemented method for controlling camera exposure to augment a wiper system of a sensor enclosure. The computer implemented method can detect a presence of a wiper in one or more images captured by one or more cameras. An exposure time of the one or more cameras can be adjusted. Wiper speed can be adjusted such that wipers move in and out of one or more field of views of the one or more cameras while the one or more cameras are capturing images.

Method for controlling camera exposure to augment a wiper system of a sensor enclosure

A cover defining an outer contour of a sensor enclosure with at least three identifiable portions stacked along a vertical axis. A first portion having a circular domed shape. A second portion, disposed underneath the first portion and coupled to a base of the first portion, having a truncated cone shape. The second portion includes one or more protruding grooves arranged diagonally about the vertical axis and imprinted on an outer surface of the second portion. The one or more protruding grooves channel a portion of an inlet airflow drawn into a cavity of the cover into a circular airflow. A third portion, disposed underneath the second portion and coupled to a base of the second portion, having a truncated cone shape.

Growing community of inventors

Los Gatos, CA, United States of America

Yubo Zhang