ROBOT NAVIGATION IN CONTEXT OF OBSTACLE TRAFFIC INCLUDING MOVEMENT OF GROUPS

Number: 20190176333
Published: 2019-06-13
Filed: 2017-12-13
Assignee: DISNEY ENTERPRISES, INC.
Inventors: Hager, IV; Joseph George, Honeck; Michael R., Mika; Jeremy Andrew
CPC: G01C21/3833; B25J9/1664; B25J9/1676; G01C21/20; G01C21/3807; G05D1/0219; G06T17/05
Verdict: High Hardware
Source: Google Patents · FreePatentsOnline

The keeper's note

Robot navigation in context of obstacle traffic/group movement (droid, Hager).

Abstract

A system for improving navigation of robots in a space with a plurality of pedestrians or other movable objects or obstacles. The system includes a traffic analysis assembly that has a traffic sensor(s) sensing movement of the obstacles in the space. The traffic analysis assembly further includes a processor running a flow module that processes (such as the Gunnar-Farneback optical flow algorithm) output from the traffic sensor to generate traffic analysis results, which include density values for the obstacles in the space and motion information for the obstacles in the space (e.g., speed and direction). The system includes a robot with a controller running a navigation module selecting a navigation route between a current location of the robot and a target location in the space using the traffic analysis result. The workspace is configured such that the obstacles such as pedestrians have unregulated flow patterns in the space.

Background

BACKGROUND1. Field of the Description

The present invention relates, in general, to methods and systems for controlling a robot (such as a service or entertainment robot) for safe and efficient navigation in spaces (or “workspaces”) with one to many people, either walking or operating bicycles, wheelchairs, mobility scooters, skates, Segway or similar vehicles, and the like (all may be referred to as “obstacles” or “moving objects” or the like), and more particularly, to a robot controller (or assembly for controlling a robot) that is adapted, when navigating a robot through a crowded workspace, to provide efficient control (e.g., the quickest path) in the context of pedestrian traffic in the workspace including choosing between one or more end points/choices (e.g., a targeted second location from a current or first location (or “origin”) of the robot) and/or one or more paths to a chosen end point (or “target” or “goal”).2. Relevant Background

Today, robots are widely used in a wide variety of environments including public environments where the robots may need to interact with or at least safely navigate in the presence of people. Human behavior is complex, which makes it difficult to accurately predict what any particular person will do in the future. This is a primary challenge or difficulty in planning robot interactions with people and for navigating a robot through a space (or workspace) that may include a few people or may be even be very crowded with peop

Claims

1. A system for improving navigation of robots in a space with a plurality of moving or movable obstacles, comprising: a traffic analysis assembly including a traffic sensor sensing movement of obstacles in the workspace, wherein the traffic analysis assembly further includes a processor running a flow module that processes output from the traffic sensor to generate traffic analysis results including density values for the obstacles in the space and motion information for the obstacles in the space; and in the space, a robot comprising a controller running a navigation module to select a navigation route between a current location of the robot and a target location in the workspace based on the traffic analysis results. | 11. A system for improving navigation of robots in a space with a plurality of obstacles, the space being configured such that the obstacles have unregulated flow patterns in at least one area of the space, comprising: a traffic analysis assembly including a digital camera capturing a sequence of video frames of the space including the obstacles, wherein the traffic analysis assembly further includes a flow module processing the sequence of video frames to generate traffic analysis results including motion information for the obstacles in the space; and an autonomous agent comprising a navigation module selecting, based on the traffic analysis results, a target location from a set of potential target location and a route from a current location of the autonomous agent to the selected target location in the space. | 18. A system for improving navigation of autonomous agents in a workspace open to a plurality of movable objects, comprising: a traffic sensor sensing movement of the movable objects in the workspace; a processor running a flow module that processes output from the traffic sensor to generate traffic analysis results including density values of the movable objects in the workspace and motion information for the movable objects in the workspace; an autonomous agent in the workspace; a controller running a navigation module to select a navigation route between a current location of the autonomous agent and a target location in the workspace based on the traffic analysis results.