Ego-Vehicle Visibility Risk Planning for Hidden-Agent Collisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to effectively address collisions between ego-agents and other agents due to limited visibility and miscommunication, leading to accidents.
Innovation Solution
A method for estimating the visibility area and state of other agents, computing collision risk, and planning behaviors to minimize risk, with options for informing or controlling the ego-agent to avoid collisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the system estimates visibility area and computes collision risk for all other agents, then collision risk reduction is improved, but computational complexity and processing time increase
Solution Approach 1:
The system uses visibility area estimation as an intermediary representation to simplify collision risk computation. Instead of directly computing complex collision probabilities, the system first determines whether other agents are within the ego-agent's visibility area, using this intermediate information to guide further risk assessment and behavior planning.
Solution Approach 2:
The system performs preliminary visibility area estimation and visibility state computation before final collision risk assessment and behavior planning. By pre-computing which agents are visible or invisible to the ego-agent, the system reduces the computational burden of subsequent collision risk calculations and behavior generation.
2Reliability
If the system provides detailed visibility information and collision risk warnings, then safety awareness is improved, but information overload and driver distraction may occur
Solution Approach 1:
The system applies different levels of information presentation based on local conditions - specifically, whether each detected agent is within or outside the visibility area. Agents outside the visibility area receive special attention and warning, while visible agents receive standard processing, creating a differentiated information presentation that reduces overall driver workload while maintaining safety awareness.
Solution Approach 2:
The system provides feedback to the driver about the visibility states of other agents, particularly when agents are undetected or have limited visibility. This feedback loop allows the driver to adjust their awareness and behavior based on the system's analysis, improving safety without requiring constant driver attention to all system outputs.
Data Source
AI summary
The disclosure relates to a computer-implemented method for assisting an ego-agent, the method comprising: estimating a visibility area of another agent being present in an environment of the ego-agent; computing a visibility state of the other agent with regard to the ego-agent using the estimated visibility area of the other agent; estimating a collision risk between the ego-agent and the other agent using the computed visibility state of the other agent; planning a behavior of the ego-agent by minimizing a total cost for the behavior, wherein the total cost comprise the estimated collision risk; and performing at least one of: informing the ego-agent on the estimated collision risk and/or the planned behavior of the ego-agent; outputting, dependent on the estimated collision risk, a warning that the other agent is not aware of the ego-agent; and controlling the ego-agent using the estimated collision risk and/or the planned behavior of the ego-agent.