Vehicle Occupancy Clustering Using Proximity and Sensor Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous driving (AD) vehicles face challenges in accurately detecting and classifying occupancy within vehicles, particularly in complex environments like bumper-to-bumper traffic, where existing technologies struggle to differentiate between passengers and vehicle navigation systems, leading to reduced accuracy and confusion.
Innovation Solution
The Occupancy Detection Algorithm (ODA) leverages existing technologies such as Google Maps and smartphone sensors to link GPS, accelerometer data, and proximity information to classify clusters of individuals and vehicles, using a machine learning-based engine that correlates occupants with vehicles and determines vehicle types, including autonomous vehicles with no passengers, by analyzing acceleration patterns and electromagnetic fields.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing technologies are used to detect occupancy in vehicles, then the system can identify occupants, but the accuracy is reduced and confusion occurs between passengers and vehicle navigation systems in complex environments
Solution Approach 1:
The patent segments the occupancy detection task into multiple specialized modules: a vehicle detection module that identifies vehicles using GPS and accelerometer data, a device detection module that identifies smartphones using similar sensors, and a proximity detection module that calculates distances between detected objects. This segmentation allows each module to specialize in detecting specific object types, improving overall detection accuracy while managing system complexity through modular design.
Solution Approach 2:
The patent introduces an intermediary proximity detection system that calculates spatial relationships between vehicles and smartphones. By using GPS coordinates and accelerometer data to determine proximity, the system creates an intermediate layer of analysis that helps distinguish between actual occupants (smartphones near vehicles) and false positives (navigation systems or distant devices), thereby improving occupancy detection accuracy.
2Measurement precision
If multiple sensors and data sources are integrated to improve occupancy detection accuracy, then measurement precision improves, but the difficulty of detecting and measuring increases
Solution Approach 1:
The patent divides the complex data processing task into separate detection modules, each handling specific sensor data types. The vehicle detection module processes GPS and accelerometer data from vehicles, while the device detection module processes similar data from smartphones. This segmentation reduces the difficulty of detecting and measuring by breaking down the complex multi-sensor integration into manageable, specialized sub-tasks.
Solution Approach 2:
The patent implements feedback mechanisms where detection results from one module inform the operation of other modules. For example, the vehicle detection module's output feeds into the proximity detection module, which then informs the device detection module about potential occupant devices. This feedback loop allows the system to refine occupancy detection accuracy by continuously adjusting detection parameters based on intermediate results, making the overall complex measurement process more manageable.
Data Source
AI summary
The disclosure generally relates to autonomous or semi-autonomous driving vehicles. Specifically, an embodiment of the disclosure relates to occupancy detection algorithm to classify groups of individuals having a relationship to each other in certain environments, for example, inside a vehicle. In an exemplary embodiment, the disclosure provides an apparatus to associate one or more vehicles with one or more occupants, the apparatus including: a processing system to categorize data received from a plurality of external communication devices; a detector module configured to identify a first vehicle data and a first device data from categorized data; a proximity detector configured to communicate with the vehicle detector and the device detector; and a correlation engine to receive proximity estimate and to correlate the first device with a first occupant of the first vehicle.