Usage-Based Data Volatility Measurement for Risk Assessment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for analyzing usage-based data from physical systems, such as automobiles, fail to accurately differentiate between varying operational behaviors, leading to inadequate risk assessment and pricing in industries like insurance, as they treat all line segments equally without considering initial and final speeds or environmental conditions.

Innovation Solution

A novel method using a generalized line length formula (Equation 3) that adjusts line lengths based on initial and final speeds, and incorporates additional parameters like time of day, weather, and geographic location, along with slope and area under the curve metrics, to provide a more nuanced behavioral score.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional data collection methods are used, then data can be collected from physical systems, but the data cannot accurately differentiate between varying operational behaviors

Engineering Contradiction:
Improvebehavioral differentiation accuracyVSAvoidoperational context information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent extends traditional single-value data points into multi-dimensional segments by incorporating time, speed, acceleration, and location dimensions. Each data point becomes a segment with multiple attributes, enabling nuanced behavioral differentiation through geometric analysis of these multi-dimensional trajectories.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent divides continuous operational data into discrete segments based on temporal or contextual boundaries. Each segment represents a distinct operational episode with defined start and end points, allowing individual analysis of different behavioral patterns rather than treating all data uniformly.

Inventive Principle:
Principle #1Segmentation

2Measurement precision

If all line segments are treated equally in analysis, then calculation is simple, but risk assessment accuracy deteriorates

Engineering Contradiction:
Improverisk assessment accuracyVSAvoidanalysis method complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies different analytical weights and methodologies to different segments based on their local characteristics. High-risk segments (e.g., rapid acceleration, extreme speeds) receive greater analytical emphasis and different weighting than normal operational segments, creating a non-uniform analysis approach that improves risk detection.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent dynamically adjusts analysis parameters such as segment weighting, threshold values, and comparison criteria based on the specific characteristics of each segment. This allows the system to adapt its complexity level to the risk level being assessed, applying more sophisticated analysis only where necessary.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If detailed operational parameters are collected, then behavioral analysis can be improved, but data processing complexity increases

Engineering Contradiction:
Improvebehavioral score accuracyVSAvoiddata processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts and isolates the most diagnostically valuable parameters from the full set of collected data for specific analysis purposes. Rather than processing all available data equally, the system identifies and extracts key behavioral indicators (such as acceleration patterns, speed variations, temporal patterns) that are most relevant to risk assessment.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies analysis at different levels of detail depending on the specific assessment need. For routine monitoring, a simplified analysis of key parameters is used, while for detailed risk assessment, the full multi-parameter analysis is applied to relevant segments, avoiding unnecessary processing complexity for low-risk cases.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS9082072B1Method for applying usage based data
Publication Date: 2015.07.14 WEDDING JR DONALD K
  • US9082072B1 patent drawing
  • US9082072B1 patent drawing
  • US9082072B1 patent drawing

AI summary

A method for measuring volatility of usage based data. Usage based data are measureable parameters of a system and parameters in which the system operates in order to infer the manner in which the system is operated. The method allows for inferences regarding the operator or operators of the system and the environment in which it operates. This invention is described with reference to consumer automobiles and drivers in typical environments, primarily for application in the automobile insurance industry where measuring volatility to determine behavior. However, this method may be applied to other systems including commercial automobiles and trucks, motorcycles, all terrain automobiles, snow mobiles, water craft, air craft, space craft, assembly line operation, robotic devices, remote operational devices, and military automobiles. This method may also be applied to applications other than assessing risk including, but not limited to, marketing, profiling, sales, optimization, operations and control.