Vehicle Anomaly Detection Using Grid-Based Local Evaluation Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional anomaly detection devices in vehicles suffer from decreased accuracy in detecting anomalies due to the use of evaluation models created from wide travel regions, leading to undetected cyberattacks or unauthorized vehicle control, especially on ordinary roads.
Innovation Solution
Anomaly detection device that utilizes evaluation models specific to local regions by imposing a grid on a map, determining anomaly levels based on vehicle information, and correcting these levels based on data counts and travel speed to improve detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If evaluation models are created from wide travel regions, then the coverage area is increased, but the detection accuracy decreases
Solution Approach 1:
The patent divides the wide travel region into multiple local regions (e.g., urban areas, rural areas, highways) and creates separate evaluation models for each region. This segmentation allows the system to maintain broad coverage while achieving high detection accuracy in each specific region by using region-specific baseline behaviors rather than a single generalized model.
Solution Approach 2:
The patent implements local quality by tailoring evaluation models to specific regional characteristics. Each local region has its own evaluation model that reflects the typical driving patterns and behaviors of that area, enabling the system to adapt to local conditions and improve detection accuracy for region-specific anomalies while maintaining overall system coverage.
2Ease of manufacture
If evaluation models are created from wide travel regions, then data collection is simplified, but anomaly detection accuracy decreases
Solution Approach 1:
The patent segments the data collection process by region, collecting and analyzing data separately for each local region. This approach maintains the simplicity of data collection (as each region can be processed independently) while improving anomaly detection accuracy by comparing against region-specific baselines rather than a generalized wide-region model.
Solution Approach 2:
The patent performs preliminary action by pre-collecting and analyzing data for each local region to establish region-specific evaluation models before actual anomaly detection begins. This preliminary regional analysis simplifies the ongoing detection process while ensuring high accuracy through localized baseline establishment.
3Measurement precision
If local region evaluation models are used, then detection accuracy is improved, but system complexity increases
Solution Approach 1:
The patent manages system complexity through segmentation by organizing multiple local evaluation models into a structured regional framework. Each region has its own model, but the overall system manages these models systematically, allowing high detection accuracy through local specialization while maintaining manageable complexity through modular organization.
Solution Approach 2:
The patent applies universality by creating a multi-functional evaluation system that can handle multiple regions with different characteristics using a common framework. The system universally applies the same anomaly detection methodology across all regions while adapting the baseline parameters to each region's specific characteristics, thereby improving accuracy without proportionally increasing complexity.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
An anomaly detection device (100) includes: an obtainer (1001) that obtains vehicle information related to the status of a vehicle (30) and including location data indicating the location of the vehicle (30); a model storage (1002) that stores, for each of a plurality of cells of a grid imposed on a map, an evaluation model for evaluating the vehicle information of the vehicle (30) located at the cell; and a determiner (1003) that calculates, based on the vehicle information and evaluation models each being the evaluation model and corresponding to evaluation cells including a first cell including the location of the vehicle (30) indicated in the location data and one or more second cells each having a predetermined positional relationship with the first cell, an anomaly level indicating a degree of anomaly of the vehicle information, determines, based on the anomaly level, whether the vehicle information is anomalous, and outputs a determination result.