Location-Synchronous Vehicle Data Averaging With KDE for Road Hazards
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems are labor-intensive and reactive in identifying road surface defects and icy conditions, lacking proactive prediction capabilities due to varied road and ambient conditions.
Innovation Solution
A system for location-synchronous averaging of connected vehicle data using cloud servers to aggregate and analyze vehicle data, employing kernel density estimation (KDE) and rasterization to identify high-risk road locations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If labor-intensive surveying and purpose-built measurement systems are used to identify road surface quality, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent uses existing vehicle sensors to detect road surface conditions instead of deploying dedicated measurement systems. Vehicles naturally carry sensors for other purposes (accelerometers, GPS, wheel speed sensors) that can be repurposed to detect road defects, ice, and weather conditions. This copying approach eliminates the need for complex purpose-built measurement systems while maintaining detection capability.
Solution Approach 2:
The system leverages the vehicles' own existing sensors and operational data to identify road conditions. Rather than requiring external surveying equipment, the vehicles themselves generate the measurement data through their normal operation, making the system self-sufficient and eliminating dedicated measurement infrastructure.
2Reliability
If traditional reactive vehicle features are used to manage vehicle stability, then ease of operation is maintained, but reliability deteriorates due to lack of proactive prediction
Solution Approach 1:
The system performs preliminary detection of road conditions by aggregating data from multiple vehicles before individual vehicles encounter problematic sections. By continuously collecting and analyzing sensor data from the fleet, the system can predict upcoming road defects, ice patches, or weather changes and alert vehicles in advance, allowing proactive rather than reactive stability management.
Solution Approach 2:
The system creates a feedback loop where data from vehicles experiencing road conditions is aggregated and processed to generate predictions that are fed back to other vehicles. This collective learning mechanism allows the fleet to improve its predictive capability over time, with each vehicle benefiting from the experiences of others, thereby enhancing reliability through continuous improvement.
3Measurement precision
If individual vehicle sensor data is used alone, then device complexity is minimized, but measurement precision deteriorates due to varied road and ambient conditions
Solution Approach 1:
The patent merges data from multiple vehicles traveling on or near the same road sections to compensate for individual vehicle limitations. By aggregating sensor readings, GPS locations, and environmental data across the fleet, the system creates a more comprehensive and accurate picture of road conditions than any single vehicle could provide alone, thereby improving measurement precision through collective data fusion.
Solution Approach 2:
The system segments the road network into discrete sections and aggregates data specifically for each segment. By grouping vehicle data by geographic location and road segment, the system can analyze conditions for specific sections independently, allowing precise identification of problematic areas while accounting for local variations in road and ambient conditions.
Data Source
AI summary
Location-synchronous averaging of connected vehicle data is provided. Connected vehicle data, including events occurring to a plurality of vehicles driving along roadways and location coordinates at which the respective events occurred, is aggregated. A kernel density estimation is performed to smoothly aggregate neighboring event counts. The locations of conditions along the roadways are indicated as being where the aggregate neighboring event counts have a value over a predefined threshold.