Autonomous Vehicle Collision Avoidance via Dynamic Path Modification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional autonomous vehicles lack the ability to dynamically avoid obstacles by altering their navigation path, especially when obstacles have dynamic characteristics, and cannot effectively manage collisions by modifying their path or controlling speed.
Innovation Solution
A collision avoidance device that gathers obstacle and vehicle information, computes a collision factor, and either stops the vehicle or modifies its navigation path based on the availability of alternate routes, using a processor and memory to execute instructions for obstacle detection, path planning, and actuation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional autonomous vehicles use automatic braking control to avoid obstacles, then collision avoidance is achieved, but the vehicle cannot alter its navigation path to avoid obstacles
Solution Approach 1:
The system dynamically switches between two collision avoidance strategies: automatic braking control for static or unavoidable obstacles, and navigation path modification for dynamic or avoidable obstacles. This dynamic adaptation allows the vehicle to select the most appropriate avoidance method based on real-time obstacle characteristics, resolving the contradiction between reliable collision avoidance and adaptable navigation.
Solution Approach 2:
The system changes the control parameters by computing a collision factor that incorporates relative velocity and obstacle state. When the collision factor exceeds a threshold and alternate paths are available, the system transitions from speed control (braking) to path modification, thereby adapting the avoidance strategy based on computed parameters rather than using a fixed approach.
2Reliability
If conventional autonomous vehicles rely on speed control to avoid obstacles, then collision avoidance is achieved, but the vehicle cannot overcome obstacles with dynamic characteristics
Solution Approach 1:
The system computes a collision factor that explicitly incorporates relative velocity and obstacle state parameters. This parameter-based approach enables the system to distinguish between static and dynamic obstacles, and to select appropriate avoidance strategies: speed control for low-risk scenarios and path modification for high-risk scenarios involving dynamic obstacles, thereby overcoming the limitations of conventional speed-only control.
Solution Approach 2:
The system dynamically adjusts its avoidance strategy based on real-time computation of collision factors that consider obstacle dynamics. When obstacles exhibit dynamic characteristics (movement, changing position), the system transitions from static speed control to dynamic path modification, enabling effective handling of diverse obstacle types that conventional systems cannot address.
3Reliability
If the autonomous vehicle stops before exhausting collision distance, then passenger safety is ensured, but the vehicle cannot efficiently navigate around obstacles when alternate paths are available
Solution Approach 1:
The system uses the collision factor as a decision parameter that incorporates both safety considerations (relative velocity, collision distance) and navigational efficiency (availability of alternate paths). When the collision factor is high and alternate paths exist, the system prioritizes efficient path modification over conservative stopping, thereby maintaining safety while improving navigation productivity through intelligent parameter-based decision-making.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Method of controlling an autonomous vehicle and a collision avoidance device thereof are disclosed. The method includes gathering obstacle information associated with an obstacle around a current navigation path of the autonomous vehicle and vehicle information associated with the autonomous vehicle. The method further includes determining a relative velocity and a collision distance between the obstacle and the autonomous vehicle and an obstacle state of the obstacle based on an analysis of the obstacle information and the vehicle information. The method includes computing a collision factor based on the relative velocity and the obstacle state. When the collision factor is above a collision threshold, the method includes stopping the autonomous vehicle before exhausting the collision distance when at least one alternate navigation path is not available, or modifying the current navigation path of the autonomous vehicle when the at least one alternate navigation path is available.