Fiber Optic Cable Quality Prediction From Coating Line Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for monitoring the quality of loose tube fiber optic cables during manufacturing are time-consuming and often inaccurate, leading to delayed detection of quality drops, as they rely on offline measurements that require additional resources and are not very precise.
Innovation Solution
Implementing a machine-learning algorithm, specifically neural networks or Bayesian classifiers, to predict the quality of fiber optic cables in real-time during the secondary coating process, allowing for automated and accurate monitoring of key properties like excess fiber length, tube shrinkage, and light attenuation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If offline measurements are performed periodically to monitor fiber optic cable quality, then measurement precision is improved, but loss of time increases and productivity decreases
Solution Approach 1:
The system performs preliminary measurements of production process parameters continuously during manufacturing, before quality degradation becomes apparent. These preliminary data are fed into the machine learning model that predicts quality metrics in real-time, enabling early detection and prevention of quality issues rather than waiting for periodic offline measurements
Solution Approach 2:
The patent replaces the mechanical offline measurement system with a computational system using machine learning algorithms. The ML model substitutes physical quality measurement devices by predicting quality metrics from process parameters, enabling continuous real-time monitoring without the time delays inherent in periodic offline sampling
2Measurement precision
If offline measurements are performed periodically to monitor fiber optic cable quality, then measurement precision is improved, but productivity decreases
Solution Approach 1:
The patent replaces resource-intensive offline measurement systems with a computational machine learning approach. The ML model processes production parameter data to predict quality metrics, eliminating the need for periodic stopping of production for offline measurements, thereby maintaining high productivity while achieving accurate quality assessment
Solution Approach 2:
The system uses the existing production process data and parameters that are already being collected during manufacturing. By leveraging this self-generated data stream, the system eliminates the need for separate dedicated measurement resources and operations, allowing continuous quality monitoring without impacting production throughput
3Loss of time
If online measurements are performed during manufacturing to improve productivity, then loss of time decreases, but measurement precision worsens
Solution Approach 1:
The machine learning model acts as an intermediary between production process parameters and quality metrics. Instead of directly measuring quality parameters online with imprecise devices, the system measures easy-to-obtain process parameters and uses the ML model to infer quality metrics, achieving both real-time detection and high accuracy
Solution Approach 2:
The patent replaces direct physical measurement devices with a computational inference system. The ML model substitutes for imprecise online measurement hardware by calculating quality metrics from process parameters, thereby achieving accurate real-time quality assessment without relying on flawed direct measurement equipment
4Device complexity
If periodic offline measurements are used to monitor quality, then device complexity is reduced, but loss of information increases
Solution Approach 1:
The system implements continuous monitoring of production process parameters and continuous prediction of quality metrics through the machine learning model. This continuous action ensures that quality information is constantly available and updated, preventing any loss of quality data that would occur with periodic measurements and enabling immediate detection of quality trends
Data Source
AI summary
According to an aspect, there is provided a method for monitoring quality of loose tube fiber optic cable during manufacture in a secondary coating line. Initially, a trained machine-learning algorithm for calculating expected values of one or more quality metrics of manufactured loose tube fiber optic cable based on values of the one or more production process parameters of the secondary coating line is maintained in a machine-learning database. A computing system monitors one or more values of the one or more production process parameters during miming of the secondary coating line and calculates, in real-time during the monitoring, one or more expected values of the one or more quality metrics using the trained machine-learning algorithm with the monitored values of the one or more production process parameters as input. The computing system outputs at least the one or more expected values of the one or more quality metrics to a user device.