Impact-Specific Collision Prediction to Cut False Positives
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Sophisticated collision detection systems often result in false positives due to movements resembling vehicular collisions but not being actual collisions, necessitating a distinction between different types of impacts.
Innovation Solution
A collision prediction algorithm trained on kinematic variables using a dataset of counterfactual collisions and actual historical instances to determine a likelihood of a specific type of impact, outputting a prediction when the score is below a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a sophisticated collision detection system is used to improve collision detection capability, then the collision detection accuracy is improved, but false positives increase due to similar movement patterns between collisions and non-collisions
Solution Approach 1:
The patent segments the collision detection process by creating separate collision prediction algorithms for different types of impacts (e.g., roller coaster, skiing, boating, action sports, and vehicular collisions). Each algorithm is trained on specific datasets corresponding to its impact type, allowing the system to distinguish between similar movement patterns that would otherwise be confused by a single general-purpose detector.
2Reliability
If multiple collision prediction algorithms for different impact types are implemented, then false positives are reduced, but device complexity increases
Solution Approach 1:
The patent implements a universal collision prediction system that handles multiple impact types through a common architecture. The system uses a single mobile device with sensors (accelerometer, GPS, barometer) that collects data applicable to all impact types, and a centralized processing system that routes data to appropriate specialized algorithms based on the detected impact category.
Data Source
AI summary
Implementations claimed and described herein provide systems and methods for generating a prediction for a likelihood of a collision based on a respective collision prediction algorithm for a particular type of impact. In one implementation, determining, by a collision prediction algorithm, a prediction score based on a subset of variables associated with the movement of a mobile device. The prediction score associated with a likelihood that the movement is associated with a particular type of impact associated with the first collision prediction algorithm. A prediction that the movement is not association with the first type of impact when the prediction score is below a threshold score is outputted.