Consumption Percentile Meter for Cross-Facility Utility Benchmarking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing facility management systems fail to accurately compare utility consumption across different facilities due to assumptions of normally distributed random data, which is not true for periodic data, and lack the ability to account for facility class and size differences, leading to false warnings and inability to compare utility consumption patterns.
Innovation Solution
A computer-implemented consumption percentile meter that reads utility consumption data, determines facility class, size, and expected average unit consumption rates, and uses cumulative distribution data to normalize consumption, enabling comparison of facilities within the same class and size, and outputs a consumption percentile indicating how a monitored facility's consumption compares to peers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the extreme studentized deviate method is used to classify utility data, then data can be classified as normal or anomaly, but the assumption of normally distributed random data is incorrect for periodic data, leading to false warnings
Solution Approach 1:
The patent changes the statistical parameters and distribution assumptions from normal distribution to facility-specific patterns. It transforms the approach by using facility class, size, and temporal characteristics as new parameters for comparison, replacing the flawed extreme studentized deviate method with a customized statistical model that matches actual facility data patterns.
Solution Approach 2:
The patent applies local quality by creating facility-specific statistical models rather than using a universal normal distribution assumption. Each facility is analyzed according to its own class, size, and temporal patterns, allowing the system to adapt to local characteristics and avoid false warnings caused by generic statistical assumptions.
2Productivity
If utility consumption data from different facilities is compared directly, then comparison can be made, but facility class and size differences make such comparisons meaningless
Solution Approach 1:
The patent introduces new parameters for comparison including facility class, size, and temporal characteristics. By transforming raw consumption data into facility-specific statistical models that account for these parameters, the system enables meaningful comparisons while preserving facility-specific context through customized reference patterns.
Solution Approach 2:
The patent segments facilities into different classes and sizes, creating separate statistical models for each segment. This segmentation allows comparison within homogeneous groups while maintaining the information about facility-specific characteristics through the segmentation structure itself.
3Ease of manufacture
If a universal statistical model is used for all facilities, then implementation is simple, but it cannot account for differences in facility class and size
Solution Approach 1:
The patent creates a universal framework that handles multiple facility types through a common architecture. The system uses facility class and size as categorization parameters that enable a single system to adapt to diverse facilities, combining implementation simplicity with facility-specific customization through parameterized statistical models.
Data Source
AI summary
A statistical facility event monitor has a computer implemented event percentile meter. The event percentile meter counts the number of randomly initiated events that cause a monitored facility to consume a monitored utility over a monitored time period. The event percentile meter then calculates a cumulative distribution function for the randomly initiated events. The event percentile meter uses the cumulative distribution to determine the event percentile for the monitored facility. The event percentile meter then outputs the event percentile.