Vehicle Trip Classification Using Telematics Data

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

It is challenging for transportation network companies to accurately determine when a driver is engaged in a business trip versus personal use due to the distributed nature of their operations.

Innovation Solution

The system generates trip models based on telematics data, using features like timing and location of stops to classify trips as business or personal. Machine classifiers are trained on historical data to improve classification accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If telematics data is collected and analyzed using machine classifiers, then trip classification accuracy is improved, but system complexity increases

Engineering Contradiction:

Improvetrip classification accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system segments the trip classification problem into multiple components: telematics data collection, trip model generation, feature extraction (timing and location of stops), and machine classification. This segmentation allows each component to be optimized independently while maintaining overall system accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Trip models serve as an intermediary layer between raw telematics data and final classification decisions. These models process and structure the data, extracting meaningful features that improve classification accuracy while simplifying the machine classifier's task.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If multiple trip models are combined with driver attributes for classification, then classification reliability is improved, but data processing complexity increases

Engineering Contradiction:

Improveclassification reliabilityVSAvoiddata processing complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system merges multiple trip models with driver attributes to create a comprehensive classification framework. By combining these different data sources and models, the system achieves more reliable classification outcomes that account for both trip-specific and driver-specific patterns.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The classification system is designed to handle multiple types of inputs (different trip models and various driver attributes) through a unified machine classifier framework, making the system versatile and adaptable to different classification scenarios.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS12305994B2Systems and methods for classifying vehicle trips

Publication Date: 2025.05.20 ALLSTATE INSURANCE COMPANY

AI summary

Systems and methods in accordance with embodiments of the invention can obtain and use a variety of telematics data to classify trips taken by a vehicle. Trip models can be generated based on telematics data captured during the operation of a vehicle. A variety of features of the trip model, such as the timing and/or location of stops made by the vehicle during one or more trips, can be used to classify the trip as a business trip or a personal trip. In several embodiments, machine classifiers are trained to classify features within the trip models based on historical trips that have been classified as business trips or personal trips. A number of trip models can be combined with other driver attributes to classify a particular vehicle and/or driver as engaged with a transportation network company.