Analog Flow Meter Discrepancy Categorization via Computer Vision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Analog fluid flow meters are not connected to communication networks, making it time-consuming and costly to collect consumption data, leading to delayed detection of anomalies like leaks, which can result in resource wastage and damage.
Innovation Solution
A method using computer vision to analyze measurement readouts from analog flow meters, comparing them with historical data to categorize discrepancies as errors or anomalies by aggregating data from hierarchical levels, and issuing requests for updated readouts or anomaly events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If manual collection of consumption data from analog flow meters is used, then device complexity is reduced, but productivity deteriorates due to time-consuming and costly data collection processes
Solution Approach 1:
The patent uses computer vision to capture images of the analog flow meter display and creates a digital copy of the measurement readout. This allows automated analysis without physically interacting with the flow meter, resolving the contradiction by enabling efficient data collection while maintaining simplicity of the physical measurement device.
Solution Approach 2:
The patent replaces manual mechanical data collection with an automated optical system using computer vision and image processing. The measurement module captures images and automatically extracts consumption data, eliminating the need for manual reading and significantly improving productivity while keeping the flow meter itself simple and unchanged.
2Device complexity
If manual data collection from analog flow meters is performed, then device complexity remains low, but loss of time increases due to infrequent monitoring intervals
Solution Approach 1:
The patent enables continuous or frequent monitoring by having the measurement module automatically capture images and transmit data at regular intervals. This continuous data collection allows immediate detection of anomalies such as leaks, resolving the time loss issue while keeping the system simple through automated operations.
Solution Approach 2:
The patent implements a feedback mechanism where consumption data is continuously collected, analyzed, and compared against thresholds to detect anomalies. The system automatically identifies and reports unusual consumption patterns, enabling timely anomaly detection without increasing physical system complexity.
3Reliability
If measurement modules perform frequent repetitive readouts to detect anomalies, then reliability of anomaly detection improves, but use of energy increases reducing battery life
Solution Approach 1:
The patent applies partial action by performing repetitive readouts selectively rather than continuously. The system increases monitoring frequency only when anomalies are detected or suspected, while maintaining normal intervals during stable operation. This approach maintains anomaly detection reliability while significantly reducing overall energy consumption and extending battery life.
Solution Approach 2:
The patent implements dynamic monitoring where the readout frequency adapts based on detected conditions. The measurement module adjusts its operation from normal intervals to frequent repetitive readouts only when needed for anomaly detection, optimizing the balance between reliability and energy consumption.
4Measurement precision
If hierarchical validation with multiple measurement modules is implemented, then measurement precision improves for error categorization, but device complexity increases
Solution Approach 1:
The patent segments the validation process into hierarchical levels where individual measurement modules first perform local validation, and discrepancies are then resolved at higher levels. This segmentation allows precise error categorization while keeping each individual module simple and manageable.
Solution Approach 2:
The patent introduces a service server as an intermediary that coordinates the hierarchical validation process. The server aggregates data from multiple measurement modules, performs validation, and resolves discrepancies without requiring complex peer-to-peer communication between modules, thus improving measurement precision while managing system complexity centrally.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables remote, efficient monitoring of fluid consumption, reducing the risk of undetected anomalies and extending battery life in measurement modules by classifying discrepancies quickly and accurately.
Implementation Method 1
a digital camera unit configured to capture images of the analog numbers and/or analog pointers displayed by the analog flow meter
Implementation Method 2
obtain measurement readouts of analog numbers and/or analog pointers displayed by an analog flow meter by using computer vision
Data Source
Figure 1~2
Figure 3a
Figure 4
AI summary
The present invention relates to a method for categorizing a discrepancy in a measurement readout from a first measurement module configured to obtain measurement readouts of analog numbers and/or analog pointers displayed by an analog flow meter by using computer vision. The method comprising: comparing the measurement readout of the first measurement module with historical measurement readouts of the first measurement module thereby determining whether a discrepancy in the measurement readout of the first measurement module is present or not; and upon a discrepancy in the measurement readout of the first measurement module is present, categorizing the discrepancy as a measurement error or an anomaly by comparing a sum of a respective measurement readout from each of a first set of measurement modules belonging to a first hierarchical level in which first set the first measurement module is a member with an aggregate of a respective measurement readout from each of a second set of measurement modules belonging to a second hierarchical level, the second hierarchical level being upstream the first hierarchical level.