A machine learning method includes: a first learning step which is performed in a phase before a neural network is installed in a mobile robot and in which a stationary first obstacle is placed in a set space and the first obstacle is placed at different positions using simulation so that the neural network repeatedly learns a path from a starting point to the destination which avoids the first obstacle; and a second learning step which is performed in a phase after the neural network is installed in the mobile robot and in which, when the mobile robot recognizes a second obstacle that operates around the mobile robot in a space where the mobile robot moves, the neural network repeatedly learns a path to the destination which avoids the second obstacle every time the mobile robot recognizes the second obstacle.

Systems and methods to systematically identify and collect operational vehicle data for selective transmission to a non-local storage location for further analysis and use in training autonomous and semi-autonomous vehicles are provided. The systems and methods provided overcome limitations in storage and transmission of collected data by selectively archiving only that collected driving data warranting further analysis.

A robotic work tool system (200) for defining a working area perimeter (105) surrounding a working area (150) in which a robotic work tool (100) is intended to operate. The robotic work tool system (200) comprises a boundary definition unit (300) comprising at least one position unit (175) for receiving position data; and at least one controller (210) for controlling operation of the boundary definition unit (300). The controller (210) being configured to receive, from the position unit (175), position data while the boundary definition unit (300) is moved around the working area (150) to define a preliminary working area perimeter (110). The controller (210) is further configured to identify, based on the received position data, a geometry of the preliminary working area perimeter (110) approximately corresponding to a predefined geometry; and to adjust the identified geometry to define an adjusted working area perimeter (105), wherein the identified geometry is adjusted to correspond to the predefined geometry.

Embodiments of this specification provide a working map construction method and apparatus, a robot, and a storage medium. The method includes: determining a moving path of a robot when the robot moves forward as a first forward moving path; determining, after the robot moves backward, a position of the robot when the robot changes from moving backward to moving forward again as a correction position; determining an auxiliary position on the first forward moving path according to the correction position in a case that the correction position is not on the first forward moving path; and determining a correction path according to the correction position and the auxiliary position, so as to construct a working map of the robot according to the correction path and the first forward moving path.

A computer, including a processor and a memory, the memory including instructions to be executed by the processor to determine a first action based on inputting sensor data to a deep reinforcement learning neural network and transform the first action to one or more first commands. One or more second commands can be determined by inputting the one or more first commands to control barrier functions and transforming the one or more second commands to a second action. A reward function can be determined by comparing the second action to the first action. The one or more second commands can be output.

Generating a map associated with an environment may include collecting sensor data received from one or more vehicles and generating a set of links to align the sensor data. A mesh representation of the environment may be generated from the aligned sensor data. A system may determine a proposed link to add, a proposed link deletion, and/or a proposed link alteration, and receive a modification comprising instructions to add, delete, or modify a link. Responsive to receiving a modification, the system may re-align a window of sensor data associated with the modification. The modification and/or sensor data associated therewith may be collected as training data for a machine learning model, which may be trained to generate link modification proposals and/or determine sensor data that may be associated with a poor sensor data alignment.

Techniques to predict object behavior in an environment are discussed herein. For example, such techniques may include inputting data into a model and receiving an output from the model representing a discretized representation. The discretized representation may be associated with a probability of an object reaching a location in the environment at a future time. A vehicle computing system may determine a trajectory and a weight associated with the trajectory using the discretized representation and the probability. A vehicle, such as an autonomous vehicle, can be controlled to traverse an environment based on the trajectory and the weight output by the vehicle computing system.

A graph representation of a tactical map representing a plurality of static components of an environment of a vehicle is generated. Nodes of the graph represent static components, and edges represent relationships between multiple static components. Different edge types are used to indicate respective relationship semantics among the static components. Individual nodes are represented as having the same number and types of edges in the graph. Using the graph as input to a neural network based model, a set of results is obtained. A motion control directive based at least in part on the results is transmitted to a motion-control subsystem of the vehicle.

The present invention provides an autonomous traveling work machine that can accurately receive positioning signals from navigation satellites and autonomously travel without deviating from a traveling path, even in the case of an inclined slope. The autonomous traveling work machine includes a traveling machine body, a positioning receiver that receives positioning signals from navigation satellites, an autonomous traveling control device that performs control for autonomous traveling along traveling paths based on the positioning signals, an inclination detection unit that detects the inclination of the traveling machine body and outputs inclination angle information, an inclination angle determination unit that determines an inclination angle based on the inclination angle information, and a rotation control mechanism that rotates the positioning receiver with one or more degrees of freedom. The rotation control mechanism keeps the positioning receiver horizontal based on the inclination angle.

A method and apparatus for determining a turn-round path of a vehicle, a device and a storage medium are provided. An embodiment of the method includes: determining a starting position and a target position for the vehicle to turn round on a road; determining, based at least partially on road information associated with the road and vehicle information associated with the vehicle, a candidate turn-round path between the starting position and the target position; evaluating the feasibility of the candidate turn-round path; and determining, based on the evaluation on the feasibility, a turn-round path by which the vehicle is to turn round on the road.