AI-Based Stuck Area Classification Model Sharing for Robot Cleaners
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Robot cleaners often get stuck due to furniture or building layouts, and existing solutions fail to efficiently share information to prevent such situations effectively.
Innovation Solution
An AI apparatus that allows multiple cleaners to share information about stuck areas by training a stuck area classification model using 3D sensor and bumper sensor data, enabling one cleaner to transmit this model to another to avoid getting stuck, utilizing machine learning and neural networks for path classification and escape route determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If robot cleaners operate independently without sharing information, then each cleaner can function autonomously, but cleaners repeatedly encounter stuck areas and waste time through trial and error
Solution Approach 1:
The patent merges the experience and data from multiple independent cleaners into a shared knowledge base. Cleaners transmit their collected information about stuck areas to each other, combining individual experiences into collective intelligence that benefits the entire fleet, thereby eliminating redundant trial-and-error behaviors.
Solution Approach 2:
The system performs preliminary action by having cleaners share information about stuck areas before other cleaners encounter them. The first cleaner to discover a stuck area transmits this information to others, allowing subsequent cleaners to avoid these areas in advance and plan alternative paths, preventing time loss before it occurs.
2Reliability
If cleaners share detailed environment information, then escape from stuck areas is improved, but communication data requirements and processing complexity increase
Solution Approach 1:
The system extracts only the essential and relevant information needed for stuck area detection and avoidance. Instead of sharing complete environmental maps or all sensor data, cleaners transmit specific information about stuck areas including location coordinates, characteristics, and escape paths, reducing data volume while maintaining detection accuracy.
Solution Approach 2:
The patent uses copying by transmitting simplified representations of stuck area information between cleaners. Rather than sharing raw sensor data or complete environment models, the system creates and transmits compact data structures containing essential stuck area characteristics, making information sharing efficient and scalable.
Data Source
AI summary
An AI apparatus and an operating method are provided, the AI apparatus includes a communication interface to receive 3D sensor data and bumper sensor data from a first cleaner, a processor to generate surrounding situation map data based on the 3D sensor data and the bumper sensor data, and a learning processor to generate learning data by labeling area classification data for representing whether the surrounding situation map data corresponds to the stuck area, and to train a stuck area classification model based on the learning data. The processor transmits the trained stuck area classification model to a second cleaner through the communication interface.