Automated Driving Control Adjusting Driver Learning by Area Characteristics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional driving control devices for automated vehicles do not adequately consider the driving characteristics of the area, leading to differences in behavior between automated vehicles and surrounding vehicles, causing occupant insecurity.
Innovation Solution
A driving assistance system that learns driving characteristics during manual driving and adjusts the learning results based on the detected driving characteristics of the area, ensuring the automated vehicle behaves similarly to surrounding vehicles by using a driving assistance device with a data-for-learning storage unit, driving characteristic learning unit, area determination unit, and automated driving control execution unit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the driving control device directly applies the learning result from the driver to the automated driving control, then the automated driving reflects the driver's driving characteristics, but the automated driving vehicle behaves differently from surrounding vehicles when the driver's driving characteristics differ from the area's driving characteristics
Solution Approach 1:
The system dynamically adjusts the learning result by changing the parameters of driving behavior based on the detected area's driving characteristics. When a mismatch is detected between the driver's learned characteristics and the area's characteristics, the system modifies parameters such as acceleration patterns, braking intensity, and speed maintenance to align with local norms, thereby resolving the contradiction between personalization and social conformity.
Solution Approach 2:
The area determination unit and driving characteristic detection unit act as intermediaries between the driver's individual characteristics and the automated driving control. These intermediary components detect and analyze the area's driving characteristics, then mediate the application of learning results by adjusting them to match local conditions, preventing direct conflict between driver preferences and area norms.
2Ease of operation
If the automated driving vehicle uses learning results from manual driving, then the system personalizes the driving behavior to the driver's preferences, but the occupant feels insecure when the vehicle behavior differs from surrounding vehicles
Solution Approach 1:
The system dynamically adjusts driving behavior parameters based on the detected area characteristics. When the learned driver characteristics would result in behavior that deviates from local norms, the system modifies parameters such as following distance, acceleration rates, and speed selection to conform to area-specific patterns, thereby maintaining occupant comfort while ensuring social acceptability.
Solution Approach 2:
The system implements feedback by continuously monitoring the detected area driving characteristics and comparing them with the learned driver characteristics. This feedback loop enables the system to identify mismatches and automatically adjust the application of learning results, ensuring that personalized driving behavior remains within acceptable boundaries for the specific area, thus preventing occupant insecurity.
3Loss of information
If the driving control device learns driving operations in association with driving environment items, then the learning becomes context-aware, but the system fails to account for regional driving characteristics leading to inappropriate behavior in different areas
Solution Approach 1:
The system applies local quality by detecting and adapting to region-specific driving characteristics. Instead of using a uniform learning approach across all areas, the system identifies local patterns in driving behavior specific to each area and adjusts the application of learned characteristics accordingly, ensuring that context-aware learning remains appropriate for the local environment.
Solution Approach 2:
The system segments the driving environment into distinct areas with different driving characteristics. By dividing the operational context into region-specific segments, the system can apply different adjustment strategies for each area, allowing context-aware learning to function effectively within each local context while maintaining overall adaptability across diverse regions.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A driving assistance method of the present invention is for an automated driving vehicle that is capable of switching between manual driving by a driver and automated driving, learns a driving characteristic of the driver during the manual driving, and reflects a learning result to a driving characteristic under control of the automated driving, and the driving assistance method includes: detecting a driving characteristic of an area in which the automated driving vehicle is traveling; and adjusting the learning result according to the detected driving characteristic of the area and executing the control of the automated driving based on the adjusted learning result.