Lidar Robot Localization Using Probability Maps From New Start Positions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot systems inefficiently regenerate map data when the traveling start position changes within the same indoor space, leading to unnecessary data generation and wastage of useful map data.
Innovation Solution
A robot equipped with a Lidar sensor, motor driver, and processor that acquires sensing data to identify the traveling start position by matching probability information and position information from existing map data, allowing continuous use of existing map data even when the start position changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the robot regenerates map data when the traveling start position changes, then the robot can adapt to the new start position, but it causes unnecessary data generation and wastes existing useful map data
Solution Approach 1:
The system changes the parameter of map data validity from binary (valid/invalid) to probabilistic (validity probability), allowing the robot to determine that map data remains valid when the traveling start position changes within the same indoor space by comparing probability values against a threshold
Solution Approach 2:
The system creates a probability map as a computational copy of the indoor space, where each location has an associated probability value indicating the likelihood that map data is valid at that location, enabling efficient determination without full map regeneration
2Measurement precision
If the robot regenerates map data when the traveling start position changes, then the robot can ensure data accuracy, but it leads to redundant data generation and inefficiency
Solution Approach 1:
The system introduces probability values as a new parameter to quantify map data validity, transforming the accuracy assessment from a binary determination to a probabilistic evaluation that preserves accurate map data while avoiding unnecessary regeneration
Solution Approach 2:
The system uses feedback from probability value comparisons to determine whether map data regeneration is necessary, creating a closed-loop control mechanism that maintains accuracy while improving efficiency by only regenerating data when truly needed
3Reliability
If the robot stops using existing map data when the traveling start position changes, then the robot can avoid using outdated data, but it causes wastage of highly useful map data
Solution Approach 1:
The system changes the reliability assessment from a binary decision to a probabilistic evaluation, allowing the robot to maintain confidence in existing map data by comparing probability values against thresholds, thereby avoiding unnecessary data regeneration and energy consumption
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 reuse of existing map data by accurately identifying and updating map data only when the traveling space changes, reducing redundant data generation and improving operational efficiency.
Implementation Method 1
a light and detection ranging ("Lidar") sensor
Data Source
AI summary
A robot includes: a light and detection ranging (“Lidar”) sensor; a driving module; a memory configured to store first map data corresponding to a first traveling space; and at least one processor configured to: acquire sensing data through the Lidar sensor at a traveling start position of the robot, control the driving module to move the robot in a state in which a position corresponding to the traveling start position of the robot is not identified on the first map data based on the acquired sensing data, acquire second map data based on the sensing data acquired through the Lidar sensor while the robot is moving, identify whether a second traveling space corresponding to the second map data matches the first traveling space based on probability information included in the first map data and position information on one or more objects included in the second map data, and identify the traveling start position of the robot on the first map data based on the traveling start position of the robot on the second map data and the position information on the one or more objects in a state in which it is identified that the second traveling space matches the first traveling space.