Long Range Sensor Segmentation for Highway Collision Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle safety systems, such as adaptive cruise control (ACC), are limited in their ability to effectively prevent collisions at high velocities on highways due to restricted sensor coverage areas, leading to potential late detection of decelerating vehicles outside their target selection range, which can result in inadequate braking and increased collision risk.
Innovation Solution
A vehicle safety assist system equipped with long-range sensors that can detect vehicles beyond the ACC's target selection range, a sensor information processor to assess vehicle velocity and deceleration, and a driver warning and brake control system that provides warnings or initiates autonomous braking based on determined relative deceleration and confidence levels, even with unreliable sensor information, to pre-brake the vehicle and enhance safety.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If ACC target selection range is used to limit sensor coverage, then device complexity is reduced, but detection capability at long range is insufficient
Solution Approach 1:
The sensor coverage area is divided into two distinct zones: an inner ACC target selection range for normal cruise control operations and an outer extended sensor coverage area for long-range detection. This segmentation allows each zone to serve its specific function optimally, with the inner zone maintaining system simplicity and the outer zone enhancing collision avoidance reliability through early detection of decelerating vehicles.
2Reliability
If sensor coverage area is extended beyond ACC target selection range, then detection capability is improved, but device complexity increases
Solution Approach 1:
Different sensor coverage qualities are applied to different spatial zones. The inner ACC target selection range uses standard detection parameters optimized for normal following distances, while the outer extended coverage area uses enhanced detection parameters specifically tuned for long-range vehicle detection. This local quality differentiation improves detection capability without uniformly increasing complexity across the entire sensor system.
3Reliability
If autonomous braking is activated with low confidence sensor information, then safety is improved, but false braking increases
Solution Approach 1:
The system performs preliminary detection and evaluation of sensor information confidence before initiating autonomous braking. By assessing the confidence level of detected decelerating vehicles in advance, the system can distinguish between genuine collision risks and false detections, enabling safe autonomous braking only when confidence thresholds are met, thus improving safety without increasing false braking incidents.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A vehicle (10), a method (100) and a vehicle safety assist system (1) are provided. The vehicle safety assist system (1) comprises long range sensors (3) arranged to determine vehicles (A, B) travelling within a sensor coverage area (110) ahead of the host vehicle (10). A sensor information processor (5) is arranged to determine if the host vehicle (10) is travelling above at least one predetermined threshold velocity and if a relative deceleration of at least one determined vehicle (A, B) is above at least one predetermined threshold deceleration and to determine the strength or confidence of this determination. A driver warning and brake control system (7) is arranged to provide a driver warning or perform autonomous braking in response to instructions from the sensor information processor (5) based on the above determinations.