Care-Giving Robot Interaction Modeling for Adaptive Child Engagement

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Care-giving robots can only select interaction modes based on the emotion status of the interaction object and cannot provide a more appropriate interaction mode, limiting their ability to maintain the interest and passion of the child in learning.

Innovation Solution

A data processing method that generates a growing model capability parameter matrix for the interaction object, adjusting capability parameters based on both measured and statistical data, allowing the robot to enhance its interaction capabilities beyond those of the child, thereby providing a more engaging and effective interaction experience.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the robot interacts with the child at the same capability level, then the child feels equal and comfortable, but the child's learning interest and passion decrease over time

Engineering Contradiction:
Improveinteraction capability levelVSAvoidchild's learning interest and passion
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The robot dynamically adjusts its capability level based on the child's current level, transitioning from a static equal-level interaction to a dynamic adaptive interaction where the robot maintains a slight capability advantage to sustain the child's learning interest

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the capability parameter of the robot's interaction model from matching the child's level to exceeding it by a margin, using the growing model capability parameter matrix to quantify and adjust this parameter difference

Inventive Principle:
Principle #35Parameter changes

2Device complexity

If the robot uses only emotion status calculation for interaction mode selection, then the system complexity remains low, but the interaction effectiveness is limited

Engineering Contradiction:
Improveinteraction mode selection systemVSAvoidinteraction effectiveness
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The emotion recognition module serves multiple functions: it not only determines interaction modes but also provides input for capability parameter adjustment, making the same data source serve both emotional and capability assessment purposes

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system pre-calculates the growing model capability parameter matrix based on statistical data before actual interaction, preparing capability benchmarks in advance to guide real-time interaction decisions without adding computational burden during interaction

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3593958B1Data processing method and nursing robot device
Publication Date: 2022.01.05 HUAWEI TECH CO LTD
  • EP3593958B1 patent drawingFigure 1~2
  • EP3593958B1 patent drawingFigure 3
  • EP3593958B1 patent drawingFigure 4

AI summary

A data processing method for a care-giving robot and an apparatus are disclosed, to resolve a prior-art problem that a care-giving robot can perform selection from specified interaction modes only by calculating an emotion status of an interaction object, but cannot provide a more appropriate interaction mode for the interaction object. The method includes: receiving, by a model engine, data of a target object, and generating a growing model capability parameter matrix of the target object (S401); adjusting, by the model engine, an capability parameter adjustment value in the growing model capability parameter matrix based on an adjustment formula coefficient or based on a standard growing model capability parameter matrix, to determine an adjusted capability parameter adjustment value (S402); determining, by the model engine, whether the adjusted capability parameter adjustment value exceeds a preset threshold (S403); and sending, by the model engine, the adjusted capability parameter adjustment value to a machine learning engine if the adjusted capability parameter adjustment value is within a range of the preset threshold, where the machine learning engine provides, for an artificial intelligence module based on the capability parameter adjustment value, an capability parameter required for interacting with the target object (S404).