A computer stores dense maps generated by one or more aerial vehicles. The computer generates a global graph based on the dense maps and a sparse map. The computer stores a representation of one or more paths traversed by the one or more aerial vehicles within the global graph. The computer determines a path from an origin location to a destination location based on the global graph. The determined path enables an aerial vehicle to avoid objects.

A mobile machine movable between multiple rows of trees includes one or more sensors to output sensor data indicating a distribution of objects in a surrounding environment of the mobile machine, a storage to store environment map data indicating a distribution of trunks of the multiple rows of trees, a localization processor, and a controller to control movement of the mobile machine in accordance with a position of the mobile machine estimated by the localization processor. The localization processor is configured or programmed detect the trunks of the rows of trees in the surrounding environment of the mobile machine based on the sensor data that is repeatedly output from the one or more sensors while the mobile machine is moving, and perform matching between the detected trunks of the rows of trees and the environment map data to estimate a position of the mobile machine.

An apparatus, system and method capable of providing an autonomous mobile robot hazard detection and control system. The apparatus, system and method may include: a robot having a robot body; a plurality of sensors physically associated with the robot body, and capable of detecting a hazardous condition in an operational environment; and at least one processing system at least partially physically associated with the robot body and communicatively connected to the plurality of sensors. The at least one processing system may include non-transitory computing code which, when executed by a processor of the at least one processing system, causes to occur the steps of: mapping a navigation path for the robot to traverse; detecting the hazardous condition along the navigation path based on output from the plurality of sensors; and instructing at least one action by the robot other than following the navigation path, wherein the at least one action at least partially addresses the hazardous condition.

An autonomous environmental perception, path planning and dynamic landing method includes: obtaining three-dimensional environment information in real time; determining a global starting point and a global end point, and generating an initial path; optimizing the initial path based on a local path optimization algorithm to obtain a first optimized path; when a perception threshold of the current position of the unmanned aerial vehicle is greater than a preset threshold, optimizing the initial path based on a frontier-perceived path optimization method to obtain a second optimized path and a local end point; when the unmanned aerial vehicle advances to the local end point, switching to optimizing the initial path in real time based on the local path optimization algorithm; and when the unmanned aerial vehicle arrives at the global end point, carrying out dynamic landing based on a deep reinforcement learning algorithm.

A method for performing an autonomous inspection comprises traversing, by an autonomous sensor apparatus, a path through a site having three-dimensional objects located therein. The method comprises obtaining, by a plurality of sensors on-board the autonomous sensor apparatus, one or more data sets throughout the path. Each of the one or more data sets are associated with an attribute of one or more three-dimensional objects. The method comprises generating, by the first, second, or third processor, a working model from a collocated data set; and comparing, by the first, second, or third processor, the working model with one or more pre-existing models; to determine the presence and/or absence of anomalies. The presence and/or absence of anomalies are communicated as human-readable instructions.

Disclosed are apparatuses, systems, and techniques that train and use trained language models to assist users with complex systems installation, troubleshooting, and/or maintenance. A method can include generating, for a real environment including a real robot having one or more real sensors, a simulated environment modeling the real environment, the simulated environment including a simulated robot corresponding to the real robot, the simulated robot including one or more simulated sensors corresponding to the one or more real sensors, obtaining simulated data based at least on simulated sensor data collected using the one or more simulated sensors, and using the simulated data to control operation of the real robot within the real environment.

A computer stores dense maps generated by one or more aerial vehicles. The computer generates a global graph based on the dense maps and a sparse map. The computer stores a representation of one or more paths traversed by the one or more aerial vehicles within the global graph. The computer determines a path from an origin location to a destination location based on the global graph. The determined path enables an aerial vehicle to avoid objects.

An apparatus, system and method capable of providing an autonomous mobile robot hazard detection and control system. The apparatus, system and method may include: a robot having a robot body; a plurality of sensors physically associated with the robot body, and capable of detecting a hazardous condition in an operational environment; and at least one processing system at least partially physically associated with the robot body and communicatively connected to the plurality of sensors. The at least one processing system may include non-transitory computing code which, when executed by a processor of the at least one processing system, causes to occur the steps of: mapping a navigation path for the robot to traverse; detecting the hazardous condition along the navigation path based on output from the plurality of sensors; and instructing at least one action by the robot other than following the navigation path, wherein the at least one action at least partially addresses the hazardous condition.

A computer stores dense maps generated by one or more aerial vehicles. The computer generates a global graph based on the dense maps and a sparse map. The computer stores a representation of one or more paths traversed by the one or more aerial vehicles within the global graph. The computer determines a path from an origin location to a destination location based on the global graph. The determined path enables an aerial vehicle to avoid objects.

A computer-implemented method when executed by data processing hardware causes the data processing hardware to perform operations. The operations include detecting a candidate support surface at an elevation less than a current surface supporting a legged robot. A determination is made on whether the candidate support surface includes an area of missing terrain data within a portion of an environment surrounding the legged robot, where the area is large enough to receive a touchdown placement for a leg of the legged robot. If missing terrain data is determined, at least a portion of the area of missing terrain data is classified as a no-step region of the candidate support surface. The no-step region indicates a region where the legged robot should avoid touching down a leg of the legged robot.