Machine-Learned Position Estimation for Low-Power 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 costs.
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 and power consumption.
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 increases
Solution Approach 1:
The patent pre-calculates and stores correspondence relationships between point cloud data and image data in a database before actual position estimation is needed. During real-time operation, the system queries pre-computed results from the database rather than performing complex arithmetic operations, enabling real-time position estimation without requiring high-performance arithmetic devices, thus reducing power consumption while maintaining estimation accuracy
Solution Approach 2:
The patent creates a simplified representation (copy) of the complex point cloud-to-image correspondence relationship through pre-computed lookup tables. Instead of performing complex geometric transformations and matching operations in real-time, the system uses these pre-computed copies to quickly determine position information, reducing computational load and power consumption while preserving estimation accuracy
2Measurement precision
If high-performance arithmetic devices are used to perform self-localization estimation, then position estimation accuracy is improved, but system cost increases
Solution Approach 1:
The patent performs complex correspondence relationship calculations in advance during an offline learning phase, storing results in a database. During online position estimation, the system simply queries pre-computed results, eliminating the need for high-performance arithmetic devices while maintaining accurate position estimation through the pre-calculated correspondence relationships
Solution Approach 2:
The patent creates simplified lookup table structures that store pre-computed correspondence relationships between point cloud and image data. These simplified representations replace the need for complex real-time arithmetic operations, allowing accurate position estimation using standard hardware while reducing device complexity and cost
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.