A method and corresponding apparatus for collision-free motion planning of a first manipulator in a first working space and a second manipulator in a second working space, wherein the first and second working spaces at least partially overlap. The method includes the steps of importing a first dynamic roadmap for a first configuration space of the first manipulator, wherein the first dynamic roadmap includes a first search graph and a first mapping between the first working space and the first search graph, and importing a second dynamic roadmap for a second configuration space of the second manipulator, wherein the second dynamic roadmap includes a second search graph and a second mapping between the second working space and the second search graph. Furthermore, the motion of the first manipulator and the second manipulator are coordinated based on the first dynamic roadmap and the second dynamic roadmap.

This document describes a simulation system that simulates robots and other actors performing tasks in an area. In one aspect, a method includes obtaining a graph representing a physical area. The graph includes area nodes that represent regions of the area that are traversed by a set of actors that perform tasks in the area and terminal nodes that represent regions of the facility where the actors perform the tasks. A set of agents that each include a model corresponding to an actor is identified. At least a portion of the agents includes models for robots that perform tasks in the area. The model of an agent represents durations of time for traversing area nodes and performing tasks are terminal nodes during simulations. A sequence of tasks being performed in the area is simulated using the graph and the set of agents.

A computer-implemented method for planning an optimized motion path. The optimized motion path is determined applying a sampling-based motion-planning algorithm depending on a sample node set including sample nodes. The sample nodes in the sample node set are deterministically selected from a configuration node set including all obstacle free nodes. The sample nodes are selected to optimize a given dispersion criterion. The dispersion criterion selects the sample nodes so that the largest uncovered area/space within the configuration node set is as small as possible.

Collision detection useful in motion planning for robotics advantageously employs data structure representations of robots, persistent obstacles and transient obstacles in an environment in which a robot will operate. Data structures may take the form of hierarchical data structures ((e.g., octrees, sets of volumes or boxes (e.g. a tree of axis-aligned bounding boxes (AABBs), a tree of oriented (not axis-aligned) bounding boxes, or a tree of spheres)) or non-hierarchical data structures (e.g., Euclidean Distance Fields) Such can result in computational efficiency, reduce memory requirements, and lower power consumption. The collision detection can take the form as a standalone function, providing a Boolean result that can be employed in executing any of a variety of different motion planning algorithms.

The present disclosure proposes a method and an apparatus for generating an action sequence of a robot. The method includes: obtaining a directed graph, in which the directed graph comprises a plurality of nodes for instructing actions of the robot, and directed edges connecting the nodes; obtaining target actions involved in a task, and an execution order of the target actions; in the directed graph, performing a search in directions indicated by the directed edges to obtain a target path, in which nodes on the target path comprises target nodes corresponding to the target actions, and an order of the target path passing through the target nodes matches an execution order of the target actions; and generating the action sequence of the robot according to actions instructed by the target path and an execution order of the actions instructed by the target path.

A social networking application provides for automated link and/or content recommendation to users of a social media platform by automated social graph refinement that augments a baseline social graph with predicted links and inferred labels by iteratively (a) propagating attribute labels through optimizing attribute label similarity between user nodes constrained by closeness of links between the users, and (b) predicting links between users through optimizing link closeness constrained by label similarity. Each label inference iteration is based on predicted labels generated in and immediately prior link prediction iteration, and each link prediction iteration is based on inferred labels generated in and immediately prior label inference iteration.

This document describes a simulation system that simulates robots and other actors performing tasks in an area. In one aspect, a method includes obtaining a graph representing a physical area. The graph includes area nodes that represent regions of the area that are traversed by a set of actors that perform tasks in the area and terminal nodes that represent regions of the facility where the actors perform the tasks. A set of agents that each include a model corresponding to an actor is identified. At least a portion of the agents includes models for robots that perform tasks in the area. The model of an agent represents durations of time for traversing area nodes and performing tasks are terminal nodes during simulations. A sequence of tasks being performed in the area is simulated using the graph and the set of agents.

A method and corresponding apparatus for collision-free motion planning of a first manipulator in a first working space and a second manipulator in a second working space, wherein the first and second working spaces at least partially overlap. The method includes the steps of importing a first dynamic roadmap for a first configuration space of the first manipulator, wherein the first dynamic roadmap includes a first search graph and a first mapping between the first working space and the first search graph, and importing a second dynamic roadmap for a second configuration space of the second manipulator, wherein the second dynamic roadmap includes a second search graph and a second mapping between the second working space and the second search graph. Furthermore, the motion of the first manipulator and the second manipulator are coordinated based on the first dynamic roadmap and the second dynamic roadmap.

Systems and methods are provided that introduce an improved way of producing fast and optimal motion plans by using Recurrent Neural Networks (RNN) to determine end-to-end trajectories in an iterative manner. By using an RNN in this way and offloading expensive computation towards offline learning, a network is developed that implicitly generates optimal motion plans with minimal loss in performance in a compact form. This method generates near optimal paths in a single, iterative, end-to-end roll-out that that has effectively fixed-time execution regardless of the configuration space complexity. Thus, the method results in fast, consistent, and optimal trajectories that outperform popular motion planning strategies in generating motion plans.

Collision detection useful in motion planning for robotics advantageously employs data structure representations of robots, persistent obstacles and transient obstacles in an environment in which a robot will operate. Data structures may take the form of hierarchical data structures ((e.g., octrees, sets of volumes or boxes (e.g., a tree of axis-aligned bounding boxes (AABBs), a tree of oriented (not axis-aligned) bounding boxes, or a tree of spheres)) or non-hierarchical data structures (e.g., Euclidean Distance Fields) Such can result in computational efficiency, reduce memory requirements, and lower power consumption. The collision detection can take the form as a standalone function, providing a Boolean result that can be employed in executing any of a variety of different motion planning algorithms.