Mobile Object Speed Control for Obstructed Intersection Risk Areas
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile object control systems fail to appropriately control vehicles when obstacles shield certain areas, leading to inadequate recognition and management of risk zones, particularly at intersections where trajectories of multiple vehicles intersect.
Innovation Solution
A mobile object control device that recognizes surroundings, sets a risk area based on obstacles, and adjusts speed to prevent collisions by decelerating the vehicle when approaching the risk area, with the risk area size increasing as the vehicle approaches the predicted trajectory, ensuring safe navigation through intersections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the control system uses conventional collision avoidance methods, then it can handle simple scenarios, but it fails to appropriately control vehicles when obstacles shield certain areas at intersections
Solution Approach 1:
The control device performs preliminary risk assessment by determining risk areas before the vehicle reaches them. When the ego vehicle's future trajectory intersects with another vehicle's predicted trajectory and the target area length exceeds a threshold, the system proactively identifies and marks risk areas, allowing the vehicle to prepare appropriate control actions in advance rather than reacting too late
Solution Approach 2:
The patent introduces a virtual 'risk area' as an intermediary concept between the vehicle and physical obstacles. This risk area serves as a mediator that represents potential collision zones, allowing the control system to manage uncertainty and hidden dangers caused by obstructed views without needing direct sensor input from those areas
2Reliability
If the control system sets risk areas based on predicted trajectories, then it can improve safety at intersections, but it increases computational complexity
Solution Approach 1:
The control device applies local quality by focusing computational resources only on specific regions of interest rather than processing the entire environment uniformly. Risk areas are determined locally at trajectory intersections where another vehicle is detected, with the target area length threshold serving as a local criterion to trigger detailed analysis only where necessary
Solution Approach 2:
The patent segments the environment into distinct functional zones: detected vehicle positions, predicted trajectory paths, target areas, and risk areas. This segmentation allows the control system to process each element separately and systematically, reducing overall computational complexity by breaking down the complex intersection safety problem into manageable discrete components
3Reliability
If the vehicle decelerates when approaching risk areas, then collision avoidance is improved, but travel time increases
Solution Approach 1:
The control device implements dynamic speed adjustment based on the vehicle's real-time position relative to risk areas. Rather than using fixed deceleration profiles, the system continuously adapts the deceleration magnitude and timing according to the distance to the risk area boundary and the vehicle's current speed, allowing for more efficient and context-appropriate speed management
Data Source
AI summary
A mobile object control device includes a recognizer that recognizes a surroundings situation of a mobile object, and a controller that controls an acceleration and deceleration of the mobile object based on the surroundings situation recognized by recognize, and the controller sets a risk area at a first reference position based on an end of an obstacle present around the mobile object in a traveling direction of the mobile object when a predicted trajectory in which another mobile object is estimated to move and a future trajectory of the mobile object interfere with each other, and the length of a target area corresponding to the predicted trajectory in a predicted trajectory direction is equal to or larger than a predetermined length with respect to the target area difficult for the recognizer to recognize due to the obstacle, and controls at least the speed of the mobile object based on the set risk area.