Robot Gaze-Plane Control for Accurate Target Search

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current robot control systems fail to accurately determine the user's gaze direction, leading to inefficient target acquisition when users provide vague instructions, resulting in a wide search range and low accuracy.

Innovation Solution

Establishing a reference coordinate system, capturing the user's gaze direction, and calculating a gaze plane based on the robot's sight line angle and position, allowing the robot to smoothly scan the plane to find the indicated target.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the robot uses voice identification to acquire user instructions, then the robot can understand basic commands, but the robot fails to accurately determine the user's gaze direction and indicated target

Engineering Contradiction:
Improvetarget identification accuracyVSAvoidgaze direction information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent combines voice identification with gaze direction detection to create a hybrid input system. The robot integrates audio signal processing with visual attention tracking, merging two different modalities (hearing and looking) to comprehensively determine user intent and accurately identify indicated targets.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces a gaze plane as an intermediary computational construct that bridges the robot's visual field and the user's indicated target. This virtual plane serves as a mediator to translate gaze direction data into actionable target identification, resolving the information loss between user intent and robot understanding.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If the robot searches for the indicated target without gaze direction information, then the robot can operate with simple voice commands, but the search range becomes wide and search accuracy decreases

Engineering Contradiction:
Improvesearch accuracyVSAvoidsearch range
Core Design Contradiction:
Measurement precisionVSArea of stationary object

Solution Approach 1:

The patent applies local quality by concentrating the search effort within the gaze plane rather than conducting a broad omnidirectional search. By localizing the search to the specific region where the user is looking, the robot improves search accuracy while reducing the effective search area from the entire environment to a focused plane in the user's visual field.

Inventive Principle:
Principle #3Local quality

3Reliability

If the robot calculates the gaze plane in real time, then the robot can accurately track the user's indicated target, but the computational complexity increases

Engineering Contradiction:
Improvetarget acquisition reliabilityVSAvoidcontrol system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent performs preliminary actions by pre-establishing the reference coordinate system and pre-defining the gaze plane calculation methodology before actual target acquisition begins. This preparation work is done in advance so that during real-time operation, the robot only needs to execute the calculation using stored formulas and current sensor data, reducing real-time computational complexity while maintaining high reliability.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS11325255B2Method for controlling robot and robot device
Publication Date: 2022.05.10 CHONGQING XINGJIE SHUXING TECHNOLOGY PARTNERSHIP ENTERPRISE (LLP)
  • US11325255B2 patent drawing
  • US11325255B2 patent drawing
  • US11325255B2 patent drawing

AI summary

A method for controlling a robot includes: establishing a reference coordinate system; capturing a user's gaze direction of an indicated target; acquiring a sight line angle of the robot; acquiring a position of the robot; acquiring a linear distance between the robot and the user; calculating in real time a gaze plane in a user's gaze direction relative to the reference coordinate system based on the sight line angle of the robot, the position of the robot and the liner distance between the robot and the user; and smoothly scanning the gaze plane by the robot to search for the indicated target in the user's gaze direction.