Driver Input Classification for Sudden Acceleration Prevention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traffic accidents occur due to driver mistakes or carelessness, especially in novice or elderly drivers, which conventional autonomous driving technologies fail to adequately address.
Innovation Solution
A vehicle control system that limits a driver's acceleration input while allowing deceleration and steering control, incorporating autonomous driving when necessary, and includes an input interface to receive driver inputs, a processor to classify and control these inputs, and modules for environment and behavior detection to manage driving intentions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If autonomous driving technology is applied to commercial vehicles, then accidents caused by autonomous system limitations can be reduced, but accidents due to driver mistakes or carelessness cannot be adequately prevented
Solution Approach 1:
The patent introduces an intermediary system that sits between the driver and the vehicle control system. This intermediary layer classifies driver inputs into different types (emergency vs. non-emergency) and selectively processes them, allowing the system to intervene only when necessary while preserving normal driver control. This resolves the contradiction by adding a mediating layer that enhances reliability without completely removing driver adaptability.
Solution Approach 2:
The system dynamically changes the parameter of driver control authority based on the classification of driver inputs. For emergency inputs, full authority is granted; for non-emergency inputs, authority is limited or overridden by the autonomous system. This dynamic parameter adjustment allows the system to adapt reliability measures to specific driving situations, preventing accidents while maintaining necessary driver control.
2Reliability
If the system classifies and selectively incorporates driver control inputs into vehicle control, then driver mistakes can be prevented, but driver control freedom is reduced
Solution Approach 1:
The patent segments driver control inputs into different categories (emergency inputs vs. non-emergency inputs) and applies different processing rules to each segment. Emergency inputs such as sudden steering corrections or brake applications are directly transmitted to the vehicle control system, while non-emergency inputs are subject to autonomous system oversight. This segmentation allows the system to maintain driver freedom for critical maneuvers while preventing mistakes in routine operations.
Solution Approach 2:
The system applies partial control limitation only to specific types of driver inputs rather than completely restricting driver control. By selectively incorporating only certain driver inputs into vehicle control while filtering others, the system achieves sufficient safety improvement without excessively reducing driver freedom. This partial action approach balances safety requirements with operational freedom.
3Measurement precision
If the system performs learning and updates driver's driving intention detection logic, then detection accuracy improves, but system complexity increases
Solution Approach 1:
The patent implements a feedback mechanism where the system continuously learns from the comparison between predicted driving paths and actual driver intentions. The detection logic is updated based on this feedback, allowing the system to improve accuracy over time. This feedback-based learning approach enables gradual improvement of detection precision without requiring complete system redesign, thus managing complexity while enhancing accuracy.
Data Source
AI summary
Disclosed are a vehicle control system that limits a driver's driving behavior and a vehicle control method using the same. The vehicle control system that limits a driver's driving behavior includes an input interface device configured to receive a driver's control input, memory in which a program that recognizes the driver's control input and that determines whether to incorporate the results of the recognition into vehicle control has been stored, and a processor configured to execute the program. The processor controls the driver's control input, which belongs to first classification, to be not incorporated into the vehicle control without any change and controls the driver's control input, which belongs to second classification, to be immediately incorporated into the vehicle control based on the results of the recognition.