Driving Assistance Prediction Reliability via Dual Subsystem Merging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current driving assistance systems face challenges in reliability due to sensor data errors, leading to less reliable predictions and potentially inappropriate vehicle control actions, which can be uncomfortable for drivers and other traffic participants.
Innovation Solution
A method that combines predictions from two distinct prediction subsystems, one relying on indirect indicators and the other on direct indicators, to determine a control signal for active vehicle control, allowing for improved reliability and reduced error situations without the need for additional sensor equipment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a single prediction subsystem is used for vehicle control predictions, then the device complexity is reduced, but the reliability of predictions decreases due to sensor data errors
Solution Approach 1:
The prediction subsystem is divided into two independent segments: a first prediction subsystem using indirect indicators and a second prediction subsystem using direct indicators. Each subsystem operates independently to generate predictions, and their results are combined to improve overall prediction reliability while maintaining manageable complexity in each individual subsystem.
Solution Approach 2:
The predictions from the first prediction subsystem (indirect indicators) and the second prediction subsystem (direct indicators) are merged through a combination unit. This merging process integrates the strengths of both subsystems to produce a more reliable control signal that overcomes the limitations of individual sensor data sources.
2Measurement precision
If additional sensor equipment is added to improve prediction accuracy, then measurement precision increases, but device complexity and costs increase
Solution Approach 1:
The existing sensor equipment is used for multiple purposes: the first prediction subsystem processes indirect indicators for early warning predictions, while the second prediction subsystem processes direct indicators for confirmation predictions. This multi-functionality approach maximizes the utilization of existing sensors without requiring additional hardware investments.
Solution Approach 2:
The first prediction subsystem performs preliminary prediction using indirect indicators before the actual behavior occurs. This preliminary action allows the system to prepare control strategies in advance and cross-validate with the second subsystem's direct indicator predictions, improving accuracy without additional sensors.
3Reliability
If predictions are made based on limited sensor data, then the device complexity remains low, but the reliability of vehicle control actions decreases leading to driver discomfort
Solution Approach 1:
The system implements a feedback mechanism where the control signal generation process continuously compares and combines predictions from both subsystems. This feedback loop ensures that control actions are based on validated predictions from multiple sources, increasing reliability and reducing false actions that would cause driver discomfort.
Solution Approach 2:
The system uses partial actions by weighting the contributions of each prediction subsystem appropriately. Rather than requiring both predictions to be identical, the system accepts control actions based on sufficient agreement between subsystems, providing reliable control without excessive complexity while maintaining driver acceptance through appropriate action intensity.
Data Source
AI summary
The invention relates to a driving assistant adapted for active control of a vehicle based on predictions of a behavior of a detected object. A method aspect of the invention comprises accepting a first prediction of a behavior associated with the detected object from a first prediction subsystem and a second prediction from a second prediction subsystem; determining a control signal based on a combination of the first prediction and the second prediction; and initiating active control of the vehicle based on the control signal.