ADS Takeover Training Using Reliability-Based Driver Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for training drivers on automated driving systems (ADS) are time-consuming, ineffective, and fail to comprehensively educate drivers about ADS reliability and capabilities, often leading to unnecessary vehicle takeovers, which can reduce performance and safety.
Innovation Solution
A training system that generates an automated motion plan for driver takeovers, determining if the takeover was necessary by comparing it to a reliability model based on vehicle and environmental parameters, and providing feedback to the driver to improve understanding and confidence in ADS operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional training methods (manuals, virtual training systems, interactive websites) are used to train drivers about ADS operation, then drivers can obtain basic information about ADS capabilities and limits, but the training is time-consuming, inconvenient, and ineffective at comprehensively training drivers and preventing unnecessary takeovers
Solution Approach 1:
The system provides real-time feedback to drivers during actual ADS operation by analyzing takeover requests against the reliability model. When a driver requests takeover, the system evaluates whether the takeover was necessary based on current reliability parameters and provides immediate feedback explaining the system's perspective, thereby training drivers in-context without requiring separate training sessions
Solution Approach 2:
The system enables drivers to train themselves through direct interaction with the ADS during normal operation. By presenting real-world scenarios and providing feedback on their takeover decisions, the system allows drivers to learn about ADS reliability and capabilities through personal experience rather than passive consumption of training materials
2Ease of operation
If drivers are provided with comprehensive training about ADS capabilities and limits through traditional methods, then drivers may gain better understanding of when takeovers are necessary, but the training remains ineffective at preventing unnecessary takeovers due to lack of practical context
Solution Approach 1:
The system prepares drivers for potential takeovers by continuously monitoring reliability parameters and being ready to provide immediate feedback when takeovers occur. This preliminary preparation during normal operation builds driver confidence and understanding before critical situations arise, reducing unnecessary takeovers while maintaining safety
3Measurement precision
If the system evaluates each driver takeover request against the reliability model in real-time, then the system can provide accurate feedback on takeover necessity, but this requires continuous processing of reliability parameters during ADS operation
Solution Approach 1:
The system performs full reliability model evaluation and feedback provision selectively based on when drivers actually request takeover. During normal ADS operation without takeover requests, the system monitors reliability parameters but does not execute the full evaluation and feedback sequence, thereby achieving precise measurement when needed while reducing overall processing complexity
Data Source
AI summary
System, methods, and other embodiments described herein relate to improving the training of a driver during an automated driving system mode. In one embodiment, a method includes generating, in association with a maneuver by the driver that is a takeover of a vehicle, an automated motion plan associated with the maneuver. The method also includes determining if the automated motion plan is valid based, at least in part, on a condition that one or more parameters of a reliability model satisfy a threshold. The one or more parameters may be associated with the vehicle and a driving environment. The method also includes notifying the driver that the maneuver was unnecessary using a notification signal if the one or more parameters satisfy the threshold.