Fleet KPI Baseline Selection Using Bayesian Distribution Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fleet management systems face challenges in analyzing large amounts of data to accurately forecast Key Performance Indicators (KPIs) without requiring high computational power, necessitating a method to efficiently evaluate and update statistical distributions for improved monitoring and maintenance scheduling.
Innovation Solution
A method involving a computing device that receives and updates candidate statistical distributions, compares them with operations data, and selects a baseline distribution to best represent the data, triggering alerts for maintenance changes when necessary, using Bayesian and Frequentist statistics to refine and select distributions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If large amounts of data are analyzed to accurately forecast KPIs, then measurement precision is improved, but computational power requirements increase
Solution Approach 1:
The patent segments the large dataset into multiple batches and processes them sequentially using iterative Bayesian updating. Instead of loading all data at once, the system processes data in chunks, updating statistical distributions incrementally. This divides the computationally intensive task into smaller, manageable segments that can be handled with limited computational resources while still achieving accurate KPI forecasts through cumulative learning from all data points.
2Measurement precision
If multiple candidate statistical distributions are evaluated and updated, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent implements feedback mechanisms where the system continuously compares predicted KPIs against actual observed values and uses this feedback to update the statistical distributions. The Bayesian updating process incorporates new data and feedback loops to refine the candidate distributions, eliminating poor fits and strengthening accurate models. This feedback-driven approach systematically reduces the number of candidate distributions over time, managing complexity while improving precision.
Solution Approach 2:
The system dynamically adjusts the set of candidate statistical distributions based on performance feedback. Distributions that fail to accurately characterize the data are eliminated from the candidate set, while promising distributions are retained and refined. This dynamic pruning of the candidate pool reduces system complexity in later stages while maintaining high measurement precision through the evolution of the distribution set based on actual data performance.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method implemented by a computing device, of monitoring a collection of machines. The method includes receiving a plurality of candidate statistical distributions for the collection of machines. Each of the plurality of candidate statistical distributions describes a first operations data and a second operations data characterizing one or more aspects of at least one machine of the collection of machines. The method further includes updating and comparing the plurality of candidate statistical distributions based on a combination of the first operations data and the second operations data. The method further includes selecting from the plurality of candidate statistical distributions a baseline statistical distribution based on the comparing. Additionally, the method includes outputting the baseline statistical distribution, wherein the baseline statistical distribution is predicted to best describe both the first operations data and the second operations data and sending an alert to indicate the baseline statistical distribution that was discovered.