ML Position Estimation for Low-Power Real-Time Self-Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing self-localization methods in automated driving require high-performance arithmetic devices for real-time position estimation, leading to increased power consumption and cost.
Innovation Solution
A position estimation system utilizing a learning device and position estimation device that includes a comparison unit, learning unit, data acquisition unit, inference unit, and evaluation unit, employing machine learning models like convolutional neural networks to calculate parallel and rotation amounts, reducing the need for high-performance hardware.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If high-performance arithmetic devices are used to perform self-localization estimation in real time, then position estimation accuracy and real-time performance are improved, but power consumption and system cost increase
Solution Approach 1:
The patent applies preliminary action by pre-training a machine learning model using offline processing of map data and sensor data. The model learns the relationship between point cloud data and position information during this preliminary phase, so that during actual operation, the system only needs to perform lightweight inference rather than complex real-time arithmetic operations, thereby reducing power consumption while maintaining real-time performance
Solution Approach 2:
The patent replaces traditional mechanical arithmetic computation with machine learning inference. Instead of using high-performance arithmetic devices to perform complex calculations in real-time, the system substitutes this with a pre-trained neural network model that performs position estimation through pattern recognition, significantly reducing computational requirements and power consumption
2Productivity
If high-performance arithmetic devices are used to perform self-localization estimation in real time, then position estimation accuracy and real-time performance are improved, but system cost increases
Solution Approach 1:
The patent applies preliminary action by pre-training a machine learning model using offline processing of map data and sensor data. The model learns the relationship between point cloud data and position information during this preliminary phase, so that during actual operation, the system only needs to perform lightweight inference rather than complex real-time arithmetic operations, thereby reducing power consumption while maintaining real-time performance
Solution Approach 2:
The patent replaces traditional mechanical arithmetic computation with machine learning inference. Instead of using high-performance arithmetic devices to perform complex calculations in real-time, the system substitutes this with a pre-trained neural network model that performs position estimation through pattern recognition, significantly reducing computational requirements and power consumption
Data Source
AI summary
A position estimation system with low power consumption is provided. The position estimation system includes a comparison unit, a learning unit, a data acquisition unit, an inference unit, a data conversion unit, and an evaluation unit. The comparison unit has a function of calculating a first parallel movement amount and a first rotation amount on the basis of machine learning data representing geographic information. The learning unit has a function of generating a machine learning model through learning using the machine learning data, the first parallel movement amount, and the first rotation amount. The data acquisition unit has a function of acquiring acquisition data representing environmental information on the vicinity of a position estimation device. The inference unit has a function of inferring a second parallel movement amount and a second rotation amount, with use of the machine learning model, on the basis of the acquisition data and the machine learning data. The data conversion unit has a function of converting the machine learning data to evaluation data on the basis of the second parallel movement amount and the second rotation amount. The evaluation unit has a function of evaluating the degree of correspondence between the acquisition data and the evaluation data.