Social Robot User Switching Based on Gaze, Distance, and Crowd Density
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robots fail to consider user density and collect information on other users in vicinity while interacting, leading to decreased service usefulness and user inconvenience.
Innovation Solution
A robot equipped with a camera, depth sensor, and processor that collects and processes gazing and distance information to determine user interest and engagement levels, generating fusion maps to identify and interact with users based on their interest and engagement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If the robot selects a user as interaction target based only on position and distance scores, then the interaction target selection is simple, but service usefulness decreases when many users are concentrated forming a crowd
Solution Approach 1:
The patent introduces user density as a new parameter to modify the interaction target selection process. By calculating density scores based on the concentration of users in specific regions and incorporating this into the overall scoring mechanism, the system can distinguish between individual users and crowd formations, thereby improving service usefulness without excessive complexity
Solution Approach 2:
The robot performs preliminary analysis of user distribution and density before selecting an interaction target. By pre-calculating density scores for different user groups and storing this information, the robot can make more informed selection decisions that account for crowd situations, improving service effectiveness in advance
2Reliability
If the robot moves to initial position after interaction completion, then the robot returns to a safe position, but other users in vicinity are inconvenienced and must move to initial position
Solution Approach 1:
The robot collects and stores information about other users present in the vicinity during the interaction process. By pre-identifying these users and their positions, the robot can make informed decisions about whether to move to initial position or remain in place, thereby avoiding inconvenience to other users while maintaining safety
Solution Approach 2:
The patent makes the robot's post-interaction position dynamic rather than fixed. Based on real-time detection of other users in the vicinity and their interest levels, the robot can adaptively decide whether to return to initial position or stay at current location to continue serving other users, balancing safety and convenience
3Device complexity
If the robot does not collect information on other users during interaction, then the interaction process is simple, but users wanting service are inconvenienced
Solution Approach 1:
The robot continuously collects information about other users in the vicinity throughout the interaction process, rather than only before or after. This continuous monitoring allows the robot to detect users who may want service and respond appropriately, maintaining service efficiency without disrupting the current interaction
Solution Approach 2:
The patent introduces a fusion map as an intermediary data structure that integrates information about multiple users. This fusion map allows the robot to process and analyze user information efficiently, enabling service to multiple users without excessive complexity in the information collection and processing system
Data Source
AI summary
A robot is provided. The robot includes a camera, a depth sensor, a memory, and a processor configured to perform an interaction with a first user with a highest degree of interest from among a plurality of users present in vicinity of the robot, obtain gazing information of the plurality of users while performing the interaction with the first user, and obtain distance information of the plurality of users, determine an engagement level of the first user for the interaction by using gazing information and distance information of the first user from among the plurality of users, determine a degree of interest of another user by using gazing information and distance information of the first user and the another user from among the plurality of users, end the interaction with first user, and perform an interaction with the another user based on the degree of interest of the another user.