VRU Collision Prediction Using Multi-Sensor Bumper Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing collision prediction systems for vehicles struggle to accurately predict and determine the state of collisions with vulnerable road users, such as pedestrians and cyclists, due to differences in collision tendencies compared to collisions with heavier objects, and often fail to provide adequate protection in scenarios where sensor outputs are lower than normal despite a high likelihood of collision.
Innovation Solution
A collision prediction determination device equipped with surrounding area sensors (camera, radar, lidar), proximity sensors (sonar, pressure tube sensors), and interspace sensors, which acquire and analyze information to predict and determine the collision state with vulnerable road users, including revising thresholds for accurate detection and triggering protective measures like airbags and hood popup devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional collision detection systems are used, then collisions with heavy objects can be detected, but collisions with vulnerable road users (pedestrians and cyclists) cannot be accurately detected due to different collision tendencies and lower sensor outputs
Solution Approach 1:
The collision detection system is segmented into multiple independent detection modules: a first collision detection sensor (microphone) for detecting collision sounds, a second collision detection sensor (camera) for detecting collision images, and a collision determination unit. Each module targets specific aspects of collision detection, allowing the system to adapt to different collision types including vulnerable road user collisions where sensor outputs differ from conventional collisions.
Solution Approach 2:
The system changes detection parameters by using multiple sensing modalities (acoustic, visual) rather than relying on a single sensor type. The collision determination unit integrates information from different parameter domains (sound frequency, image recognition patterns) to accurately detect collisions with vulnerable road users who produce different sensor outputs compared to heavy object collisions.
2Device complexity
If a single collision detection sensor is used, then device complexity is reduced, but the ability to accurately determine collision state with vulnerable road users deteriorates
Solution Approach 1:
The system merges multiple collision detection sensors (microphone for acoustic detection, camera for visual detection) into a unified collision determination system. The collision determination unit combines information from both sensors to reliably determine collision states, particularly for vulnerable road user collisions where single-sensor detection is insufficient. This merging approach maintains manageable device complexity while significantly improving determination reliability.
3Productivity
If conventional collision detection methods are used, then general collision detection is achieved, but timely detection of serious accidents involving vulnerable road users is not achieved
Solution Approach 1:
The system performs preliminary detection actions by continuously monitoring for collision sounds through the microphone and collision images through the camera before a serious accident fully develops. The collision determination unit processes these preliminary detections to identify vulnerable road user collisions early, enabling timely detection and response. This preliminary action approach allows the system to detect serious accidents involving vulnerable road users faster than conventional methods while maintaining accurate collision type identification.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system effectively predicts and determines collision states with vulnerable road users, enhancing protection by accurately detecting collisions even when sensor outputs are lower than normal, thereby reducing the risk of injury to pedestrians and cyclists.
Implementation Method 1
surrounding area sensors (camera, radar, lidar)
Implementation Method 2
surrounding area sensors (camera, radar, lidar)
Implementation Method 3
proximity sensors (sonar, pressure tube sensors)
Data Source
AI summary
A collision prediction determination device of a subject vehicle includes: a surrounding area sensor (a camera/radar/lidar) configured to acquire surrounding area information which is information on an object present in a first monitor area including an area ahead of the subject vehicle in a traveling direction thereof; a proximity sensor (a sonar/pressure tube sensor) configured to acquire proximity information on proximity of the subject vehicle including information on a contact thereof with an object present in a second monitor area which is an area in a proximity of the subject vehicle; an interspace sensor (a sonar) configured to acquire interspace information which is information on an interspace between an in-vehicle component (a bulkhead/bumper beam) and a bumper member (a front bumper) disposed on the subject vehicle; and a collision prediction determination part configured to predict and determine a state of a collision of the subject vehicle with a vulnerable road user, including whether or not the collision has occurred, based on the surrounding area information, the interspace information, and the proximity information.