Intersection Collision Risk Assessment via Intent Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating the risk of collision at intersections are limited by measurement errors, location errors, and uncertainties in sensor data, particularly in complex intersection environments, and are computationally costly, making them unsuitable for real-time applications.
Innovation Solution
A method that determines the positions and intentions of vehicles at an intersection, estimating the need to stop based on traffic rules, without predicting potential trajectories, using a probabilistic approach to assess collision risk and adapt vehicle behavior in real-time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If physical models are used to predict vehicle trajectories, then collision risk can be estimated, but the computation time becomes too long for real-time applications
Solution Approach 1:
The patent extracts only the essential elements needed for collision risk assessment: vehicle positions, velocities, and acceleration patterns. Instead of computing complete trajectory predictions using complex physical models, the method extracts key kinematic parameters and uses them to evaluate collision risk directly, significantly reducing computation time while maintaining assessment accuracy.
Solution Approach 2:
The patent applies partial action by computing only the necessary portion of trajectory information required for collision risk assessment. Rather than predicting full vehicle trajectories over extended periods, the method calculates risk based on current kinematic states and short-term motion trends, sufficient for real-time decision-making without excessive computation.
2Reliability
If comprehensive trajectory prediction for all vehicles is performed, then collision risk assessment is more complete, but the computational cost increases significantly
Solution Approach 1:
The patent segments the collision risk assessment process into independent evaluations for each vehicle pair. Instead of computing interactions among all vehicles simultaneously, the method divides the problem into discrete pairs, assessing risk for each combination separately. This segmentation enables parallel processing and reduces the overall computational burden while maintaining comprehensive coverage of potential collision scenarios.
3Speed
If sensor data is used to determine vehicle positions and intentions, then real-time assessment is possible, but measurement and location errors introduce uncertainty
Solution Approach 1:
The patent incorporates uncertainty cushioning by evaluating a range of possible states rather than relying on single precise measurements. The method considers measurement errors and location uncertainties by assessing collision risk across multiple plausible vehicle positions and velocity states, providing a safety margin that compensates for sensor imprecision while maintaining real-time assessment capability.
Data Source
Figure 1
Figure 2~3
AI summary
The invention relates to a method for assessing the risk of collision between a first vehicle approaching an intersection and at least one second vehicle driving through said intersection, said first vehicle being provided with a measurement means and a computing means. The method according to the invention is mainly characterised in that same includes the following steps: a step of determining the respective positions of said first vehicle and each second vehicle, a step of estimating the intentions of the driver of said first vehicle and of each second vehicle, a step of estimating the need for the first vehicle and for each second vehicle to stop in accordance with the results provided by the two preceding steps and the traffic rules applicable to said intersection, and a step of estimating the risk of collision between the first vehicle and each second vehicle.