Mobile Robot Environment Learning During Charging for Object Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous mobile robots face challenges in implementing deep learning algorithms for environment mapping and object recognition due to high processing requirements, which limits their ability to operate with low system specifications and hinders efficient location-based operations.
Innovation Solution
An environment learning device for autonomous mobile robots that performs deep learning during charging cycles, utilizing a memory to store environment information and movement maps, and employing a LiDAR to recognize objects through feature point determination and convolutional neural networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If deep learning algorithms are performed during autonomous movement, then object recognition accuracy is improved, but system complexity and processing requirements increase
Solution Approach 1:
The patent performs deep learning algorithms during charging periods before autonomous movement, preparing environment maps and object recognition data in advance. This preliminary action allows the robot to have pre-processed environmental understanding without requiring intensive processing during movement, thus improving recognition accuracy while avoiding increased system complexity during operation.
Solution Approach 2:
The patent divides the deep learning processing into separate segments: environment mapping and object recognition are performed during charging periods, while autonomous movement uses the pre-processed data. This segmentation allows complex algorithms to run when power is available without compromising movement performance or requiring continuous high-power processing.
2Measurement precision
If deep learning is performed continuously during movement and charging, then recognition performance is improved, but energy consumption increases
Solution Approach 1:
The patent implements periodic deep learning execution during charging periods rather than continuous operation. The system alternates between charging/learning phases and movement/operation phases, allowing intensive processing only when power is being replenished. This periodic approach maintains recognition performance while preventing excessive energy consumption during autonomous operation.
3Adaptability or versatility
If additional processors are added for deep learning, then object recognition capability is improved, but device complexity increases
Solution Approach 1:
The patent makes the existing processor perform multiple functions: it handles both autonomous movement control and deep learning algorithms for object recognition. By utilizing the same processing unit for different tasks at different times (movement control during operation, deep learning during charging), the system achieves advanced recognition capability without adding dedicated processing hardware, thus avoiding increased device complexity.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables efficient environment mapping and object recognition with reduced computational complexity, allowing robots to operate with low system specifications and facilitating location-based operations without the need for continuous deep learning during movement.
Implementation Method 1
an external three-dimensional LiDAR to acquire distance information and reflected light information
Implementation Method 2
receives a reflected light signal reflected by an object and generates a point cloud in accordance with the reflected light signal
Data Source
AI summary
According to the present invention, disclosed are a device and a method of generating an object image, recognizing an object, and learning an environment of a mobile robot which perform a deep learning algorithm which allows a robot to create a map and load environment information acquired during the autonomous movement while the autonomous mobile robot is being charged and may be used for an application which finds out a location by finally recognizing objects such as furniture using a method of checking a location of the recognized objects to mark the location on the map.