Crowd-Aware Driving Path Planning Using VRU Cluster Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately predicting and navigating crowded environments due to the high computational demands of tracking and predicting the behavior of numerous individuals, which can reduce the acceptable range of driving paths and impose speed limitations, especially in situations like protests or high pedestrian traffic.
Innovation Solution
A system that forms clusters of vulnerable road users based on geometric shapes and collective motion, predicts VRU-blocked regions, and determines a driving path that avoids these regions, using sensors like radar and lidar to gather data and a data processing system to apply clustering metrics and predict crowd behavior.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the autonomous vehicle tracks and predicts the behavior of numerous individuals in crowded environments, then the accuracy of crowd behavior prediction is improved, but the computational load increases
Solution Approach 1:
The patent merges multiple individual VRUs into unified clusters based on spatial proximity and motion similarity. Each cluster is represented by a single geometric shape (convex hull) and collective motion parameters, reducing the number of tracked objects from numerous individuals to a manageable number of clusters while preserving essential crowd behavior characteristics.
Solution Approach 2:
The patent segments the crowd into meaningful clusters based on spatial and motion criteria. By dividing the crowded environment into discrete clusters with distinct motion patterns, the system can process each cluster independently using simplified models, reducing overall computational complexity while maintaining prediction accuracy.
2Reliability
If the autonomous vehicle narrows the acceptable range of driving paths to avoid VRU-blocked regions, then the safety of navigation is improved, but the productivity of the vehicle decreases
Solution Approach 1:
The patent dynamically adjusts the acceptable driving path range based on real-time crowd predictions. The system expands or contracts the feasible path corridor depending on the predicted positions and motions of VRU clusters, allowing the vehicle to maintain higher speeds in safer regions while automatically narrowing paths when VRU-blocked regions are predicted, thus balancing safety and productivity.
Data Source
AI summary
The disclosed systems and techniques facilitate efficient prediction of crowd behavior and safe and courteous navigation of crowded areas in driving environments. An example disclosed system includes a sensing system and a data processing system of a vehicle. The sensing system obtains sensing data associated with a driving environment of the vehicle. The data processing system detects, based on the sensing data, presence of vulnerable road users (VRUs) in the driving environment. The data processing system applies one or more clustering metrics to form cluster(s) of VRUs, each cluster associated with a geometric shape enclosing one or more VRUs and a velocity associated with collective motion of these VRUs. The data processing system predicts, using the geometric shapes and the associated velocities, one or more VRU-blocked regions and determine a driving path of the vehicle in the driving environment.