Google Inc.

Uber Technologies, Inc.

Mohammed Waleed Kadous

Andrew Lookingbill

Brian Patrick Williams

Marc Stogaitis

Joakim Kristian Olle Arfvidsson

Zhentao Sun

Sebastian Thrun

Andrew Ames Bunner

Nadav Aharony

Pascal Tom Getreuer

Alan Lee Gardner, Iii

Richard Daniel Webb

Ami Patel

Thomas Chao

Russell Heywood

Brian Duff

Sorelle Alaina Friedler

Daniel Estrada Alva

Benjamin Razon

Tony Weber

Liem Tran

Joseph A Farfel

Richard Francis Lyon

Xi Chelsea Chen

Romain Clément

Etienne Le Grand

Ryan Michael Rifkin

Scott Ettinger

Rachael Elizabeth Roberts

Srinivasan Rajagopal

Thomas Germano O'Neill

Stephen David Zelinka

Vasudev Parameswaran

Alena Rose Fong

Anup Mantri

Carl Ferman Smith

Theodore Sumers

Wei Wang

Nicholas Julian Pelly

Upamanyu Madhow

Denise Ho

Pierre B Vandwalle

Kevin Brinig

Vijay Mahadevan

Jaikumar Ganesh

Andrew T Irish

Angana Ghosh

Theodore Russell Sumers

Himaanshu Gupta

Jeff Brown

Daniel Iland

Aditya Sarawgi

Prem Kumar

Vinit Chandrakant Deshpande

Meghan Desai

Jingchen Liu

David Nicholas Christie

Venkata Sathya Praveen Gorthy

Maksim Ioffe

Isaac Richard Taylor

Timo Pekka Pylvaenaeinen

Ralph Jacob Cressman

Yuh-Jie Eunice Chen

Laurent Tu

Reto Meier

Brandon Layton

Alena Fong

A system trains and applies a machine learning model to label maps of a region. Various data modalities are combined as inputs for multiple data tiles used to characterize a region for a geographical map. Each data modality reflects sensor data captured in different ways. Some data modalities include aerial imagery, point cloud data, and location trace data. The different data modalities are captured independently and then aggregated using machine learning models to determine map labeling information about tiles in the region. Data is ingested by the system and corresponding tiles are identified. A tile is represented by a feature vector of different data types related to the various data modalities, and values from the ingested data are added to the feature vector for the tile. Models can be trained to predict characteristics of a region using these various types of input.

Determining geographical map features with multi-sensor input

A collision warning system determines probabilities of potential collisions between a vehicle and other objects such as other vehicles. In an embodiment, sensors of a client device capture sensor data including motion data and image frames from a forward-facing view of the vehicle. An orientation of the client device relative to the vehicle may be determined using the motion data. The collision warning system determines cropped portions of the image frames and detects an object captured the image frames by processing the cropped portions. The collision warning system determines a probability of a potential collision between the vehicle and the object by tracking motion of the object. Responsive to determining that the probability is greater than a threshold value, the collision warning system may provide a notification of the potential collision to a driver of the vehicle.

Image processing for vehicle collision avoidance system

An autocheck module of a map system is configured to automatically identify anomalous conditions within map data that may indicate an error within the data. The identification of the anomalous conditions is accomplished by application of different autocheck types to the map data, each autocheck type representing a class of anomalies and being triggered if particular map data exhibits the anomalous condition associated with the autocheck type. In one embodiment, for at least some of the portions of map data that trigger an autocheck type, an issue entry is created in an issue database, the issue entry referencing the autocheck type that was triggered, the map data that triggered it, and any associated data of relevance for the particular autocheck type in question.

Automated identification of anomalous map data

In context of a transportation service provided through a transport arrangement service, examples determine a point of trip termination for a trip. The point of trip termination can be determined from passive information that is collected from a computing device of a driver. The point of trip termination can be determined based on driver input and the collected information from the driver's computing device.

Trip termination determination for on-demand transport

A wearable computing device is described that detects an indication of movement associated with the wearable computing device when a user of the wearable computing device detected being located within a moving vehicle. Based at least in part on the indication of movement, a determination is made that the user of the wearable computing device is currently driving the moving vehicle. An operation is performed based on the determination that the user of the wearable computing device is currently driving the moving vehicle.

Detecting driving with a wearable computing device

Embodiments of the present disclosure support improving determination of a location of a driver device that performs bandwidth constrained communication with a server, based on sensor data acquired by the driver device. The driver device reduces dimensionality of the acquired sensor data before transmitting the sensor data to the server over a communication network. The server receives GPS data and compressed sensor data from the driver device, and determines a quality metric related to the GPS data. Based on the quality metric, the server increases dimensionality of the compressed sensor data to reconstruct original sensor data acquired by the driver device. The server than augments the GPS data with the reconstructed sensor data, and determines location information of the driver device based on the augmented data.

Locating a user device

A client device maintains location state data including a first location estimate of a geographic location of the client device. The first location estimate is based on a first motion measurement obtained over a first time period. The client device retrieves a second motion measurement obtained over a second time period subsequent to the first time period and uses it and the first location estimate to generate a second location estimate. The client device sends the second location estimate to a server. The server further processes the second location estimate to generate an updated second location estimate. The client device retrieves a third motion measurement and generates a third location estimate. The client device receives the updated second location estimate and uses it to adjust the third location estimate. A fourth location estimate is generated using the adjusted third location estimate.

Low latency client location estimation with dynamic integration of server output

A system adjusts an estimated travel time from an origin to a destination to better predict an actual trip duration. The system receives a route from a specified origin and destination. The system receives an estimated trip duration corresponding to the generated route. A machine learned model improves the estimated trip duration by applying data about past trips facilitated by the system and data about traffic-control features associated with a route. For example, the system may use counts of a number of road signs and a number of traffic signals located along the generated route to predict an actual trip duration. In some cases, the system may additionally use data about synchronized traffic lights to predict actual trip duration.

Determining adjusted trip duration using route features

The disclosure includes a system and method for detecting fine grain copresence between users. The system includes a processor and a memory storing instructions that when executed cause the system to: transmit a wakeup signal to a plurality of devices based on coarse grain location information; send a request to a first device of the plurality of devices to transmit a token using a first communication technology to determine fine grain copresence; receive a first token acknowledgment from a first subset of the plurality of devices; send a request to a second device of the first subset of the plurality of devices to transmit the token using a second communication technology to determine fine grain copresence; receive a second token acknowledgment from a second subset of the plurality of devices; and refine copresence based on receiving the first and second token acknowledgment.

Server orchestrated connectivity

Aspects of the disclosure relate generally to localizing mobile devices. In one example, a first location method associated with a first accuracy value may be used to estimate a location of the mobile device. A confidence circle indicative of a level of confidence in the estimation of the location is calculated. The confidence circle may be displayed on a mobile device. When other location methods become available, the size of the displayed confidence circle may be expanded based on information from an accelerometer of the client device or the accuracy of the other available location methods. This may be especially useful when the mobile device is transitioning between areas which are associated with different location methods that may be more or less accurate.

Position indication controls for device locations

Aspects of the disclosure relate to generating or updating an access point model. The model may be generated based in part on access point model data, e.g., collected scan information from devices that travel through an indoor space. The wireless access point model may include an indoor map of a location or building. At some point, it may be determined that the access point model data may not meet a predetermined quality threshold. For example, the data may be outdated or inaccurate. A request is generated and transmitted to a client device within a predetermined distance of the access point associated with the access point model. In response to the request, the device travels to the indoor space associated with the access point to collect scan information. The scan information is used to update or otherwise modify the access point model.

System and method for updating an access point model

Growing community of inventors

Santa Clara, CA, United States of America

Mohammed Waleed Kadous