Multi-Hypothesis Prediction Device for Autonomous Vehicle Sensor Visibility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in urban and off-highway scenarios where obstructions can prevent the detection of traffic lights or signs, leading to potential blocking scenarios due to fixed sensor positions and changing field of view.
Innovation Solution
A multi-hypothesis prediction device that analyzes sensor data and map information to determine the future field of view, allowing the vehicle to adjust its path to ensure critical points of interest, such as traffic lights, remain within the sensor's field of view, using a sensor arrangement, calculation module, and path determination module to simulate human decision-making and prevent blocking scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the vehicle moves forward along a determined path, then the vehicle progresses toward its destination, but the field of view may be obstructed by static or dynamic objects, causing loss of detection of traffic lights or signs
Solution Approach 1:
The system performs preliminary detection of potential obstructions and predicts future field of view before the vehicle reaches critical positions. By calculating the future field of view at determined future positions and identifying potential obstructions in advance, the system can adjust the path proactively to prevent loss of detection, rather than reacting after obstruction occurs
Solution Approach 2:
The system dynamically adjusts the vehicle path based on real-time detection of obstructions and predicted future field of view requirements. The path determination module modifies the planned trajectory adaptively to maintain visibility of traffic lights and signs, transforming a static path into a dynamic one that responds to changing environmental conditions
2Device complexity
If the sensor arrangement is positioned at a fixed location on the vehicle, then the device structure is simplified, but the field of view changes with vehicle movement, potentially blocking detection of critical points of interest
Solution Approach 1:
While maintaining the fixed sensor position on the vehicle, the system dynamically calculates the future field of view at multiple future positions along the determined path. This allows the system to predict visibility changes and adjust the vehicle's trajectory accordingly, maintaining detection reliability without complicating the sensor mounting structure
Solution Approach 2:
The system preliminarily determines the future field of view and identifies potential obstructions before the vehicle reaches positions where detection might be blocked. This advance calculation allows the path to be adjusted proactively to ensure the fixed sensor will maintain visibility of critical points of interest
3Reliability
If the vehicle stops to maintain visibility of traffic lights, then detection reliability is improved, but traffic flow efficiency decreases
Solution Approach 1:
The system dynamically adjusts the vehicle path to maintain visibility of traffic lights throughout movement, eliminating the need for stopping. By continuously calculating the future field of view and adjusting the trajectory in real-time, the system maintains detection reliability while keeping the vehicle in motion, thus preserving traffic flow efficiency
Solution Approach 2:
The system preliminarily identifies the optimal stopping location where the traffic light remains visible, but instead of stopping, it adjusts the path to pass through this location while maintaining visibility. This allows the vehicle to continue moving through the intersection while ensuring detection reliability
Data Source
Figure 1~3
Figure 4~5
AI summary
The invention relates to a multi hypothesis prediction device (1) for a vehicle (2). The multi hypothesis prediction device (1) comprises a sensor arrangement (11), a calculation module (10) and a path determination module (12) for determining a path (13) of the vehicle (2). The calculation module (10) is configured to determine at least one point of interest (4). The calculation module (10) is further configured to determine a field of view (3) of the sensor arrangement (11) at a future position of the vehicle (2) along the path (13). The calculation module (10) is further configured to determine, whether the point of interest (4) is within the field of view (3) of the sensor arrangement (11) at the future position, and if this is the case, to move the vehicle (2) to the future position, and if this is not the case, to not move the vehicle (2) to the future position.