Autonomous Controller Adapting to Sensor Detection Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous moving object systems face reliability issues due to varying detection accuracy of external sensors, which affects movement control, especially in environments with low detection accuracy.
Innovation Solution
A moving object controller that generates movement plans based on external and internal information, estimates detection accuracy, and selects the most reliable plan for execution, while also collecting information to improve detection accuracy when necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If the moving object relies on external sensor information for autonomous movement, then the automation level is improved, but the reliability deteriorates when detection accuracy is low
Solution Approach 1:
The system dynamically changes the control parameter (control level) based on the detection accuracy of external information. When detection accuracy is high, higher-level autonomous control is enabled; when detection accuracy is low, the control level is reduced. This resolves the contradiction by making the automation level adaptive to reliability conditions.
Solution Approach 2:
The control level is made dynamic rather than fixed. The system transitions between different control levels (fully autonomous, semi-autonomous, manual) based on real-time assessment of external information quality. This dynamic adjustment allows the system to maintain reliability while preserving automation capabilities when conditions permit.
2Reliability
If the moving object collects more external information to improve detection accuracy, then the reliability is improved, but the loss of time increases due to information collection routes
Solution Approach 1:
The system performs preliminary assessment of detection accuracy for different movement plans before execution. By evaluating which routes or actions will provide better detection accuracy opportunities, the system can proactively select plans that improve reliability without requiring time-consuming detours for data collection.
Solution Approach 2:
The moving object uses its normal movement and operation to simultaneously collect the external information it needs. Rather than dedicating separate time or routes solely for information collection, the system integrates data gathering into its regular autonomous operation, thus improving reliability without significant time penalty.
3Reliability
If the system selects movement plans with high detection accuracy, then the reliability is improved, but the adaptability worsens when no such plans exist
Solution Approach 1:
The system dynamically adjusts its plan selection criteria based on available options. When high-detection-accuracy plans are available, it prioritizes them; when only lower-accuracy plans exist, it adapts by selecting from those options while still attempting to gather information to improve future reliability. This maintains both reliability goals and adaptability to varying situational constraints.
Data Source
AI summary
The present invention provides a moving object controller and a moving object control method capable of selecting a movement plan based upon detection accuracy of external information. A moving object controller 100 includes a plan generation unit 110 that generates at least one movement plan based upon external information on an external world of a moving object 1 and internal information on an internal part of the moving object; an accuracy estimation unit 120 that estimates detection accuracy of the external information based upon the external information, the internal information, and predetermined past information, among the movement plans generated by the plan generation unit; and a plan selection unit 130 that selects a movement plan of an execution target from among the movement plans generated by the plan generation unit based upon the detection accuracy of the external information estimated by the accuracy estimation unit.