Smartphone Sensor Fusion for Accurate Vehicle Accident Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional accident detection techniques in vehicle telematics fail to leverage the abundant sensor data available on modern smartphones, resulting in inadequate resolution for effective accident response.
Innovation Solution
A method utilizing smartphone-based sensors, such as GPS and motion sensors, to collect and process movement data, supplemented with additional datasets, to accurately detect vehicular accidents and initiate appropriate responses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional vehicle telematic devices are used for accident detection, then the system has dedicated sensing capabilities, but it fails to utilize the extensive sensor data available on modern smartphones leading to inadequate detection accuracy
Solution Approach 1:
The patent applies universality by enabling smartphones to serve dual purposes: their original communication functions plus accident detection capabilities. The system utilizes existing smartphone sensors (accelerometers, GPS, gyroscopes) for telematics functions, eliminating the need for dedicated telematic devices while improving detection accuracy through multi-sensor data fusion.
Solution Approach 2:
The patent merges the functionality of dedicated vehicle telematic devices with smartphone capabilities. By combining data from multiple smartphone sensors (acceleration, position, velocity) into a unified accident detection system, the invention achieves more comprehensive data utilization and improved detection accuracy compared to traditional single-purpose devices.
2Measurement precision
If multiple sensor data sources are utilized for accident detection, then the detection accuracy improves, but the system complexity increases
Solution Approach 1:
The patent applies self-service by leveraging the smartphone's existing processing capabilities and operating system to handle sensor data fusion and accident detection algorithms. The smartphone's CPU, memory, and software stack already provide the computational infrastructure needed, eliminating the requirement for separate processing units or complex hardware architectures.
Solution Approach 2:
The system uses the smartphone's existing multi-functional architecture to handle various sensor inputs (accelerometer, GPS, gyroscope) and processing tasks (data fusion, algorithm execution, notification delivery) through a single integrated platform, thereby managing complexity while maintaining high detection accuracy.
3Loss of information
If traditional accident detection techniques are used, then the system is simple to implement, but it cannot deliver the resolution of detail needed for an ideal response to an accident
Solution Approach 1:
The patent segments the accident detection process into distinct analytical components: pre-impact detection using accelerometer data, impact phase analysis using gyroscope and accelerometer fusion, and post-impact assessment using position data. This segmentation allows detailed analysis of each accident phase while utilizing the smartphone's existing multi-sensor capabilities.
Solution Approach 2:
The patent adds temporal and spatial dimensions to accident detection by continuously monitoring sensor data before, during, and after impact events. The system analyzes acceleration patterns over time, positional changes across space, and rotational movements, creating a multi-dimensional view of the accident that provides detailed resolution for ideal response actions.
Data Source
AI summary
A method and system for detecting an accident of a vehicle, the method including: receiving a movement dataset collected at least at one of a location sensor and a motion sensor arranged within the vehicle, during a time period of movement of the vehicle, extracting a set of movement features associated with at least one of a position, a velocity, and an acceleration characterizing the movement of the vehicle during the time period, detecting a vehicular accident event from processing the set of movement features with an accident detection model, and in response to detecting the vehicular accident event, automatically initiating an accident response action.