Cable Winch Image Monitoring for Faulty Winding Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cable winch monitoring systems rely on operator attention and experience, leading to unreliable detection of winding errors, which can cause damage to cables or winches.
Innovation Solution
A device with a sensor unit and analysis unit that captures two-dimensional recordings of cable coils, using machine learning algorithms to define and analyze multiple analysis regions for detecting winding errors, including gaps, skips, and deviations, and issuing warnings or control commands.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If operator monitoring is used to detect winding errors, then the system remains simple and inexpensive, but the reliability of error detection deteriorates due to dependence on operator attention and experience
Solution Approach 1:
The patent replaces the mechanical/visual monitoring system (operator watching monitor) with an automated sensor-based system. Sensors capture images of the cable drum, and image processing algorithms automatically analyze winding behavior, substituting human visual inspection with automated optical detection and computational analysis.
Solution Approach 2:
The system performs self-monitoring and self-diagnosis by automatically capturing images, processing them through algorithms, and detecting winding errors without human intervention. The automated system serves itself by continuously monitoring its own operational state and identifying anomalies.
2Reliability
If automated sensor-based monitoring is implemented, then the reliability of winding error detection improves, but the device complexity increases due to additional sensors and processing requirements
Solution Approach 1:
The image processing is segmented into multiple analysis regions (radial edge regions, central regions, axial end regions) that are evaluated independently. This segmentation allows the system to focus computational resources on specific areas where different types of winding errors are most likely to occur, improving detection reliability while managing complexity through divided analysis.
Solution Approach 2:
Different analysis algorithms and evaluation criteria are applied to different regions of the cable drum image. Radial edge regions are analyzed for gaps and skips, while axial end regions are analyzed for stacking errors. This localized approach optimizes detection accuracy for each specific error type in its most vulnerable region.
3Measurement precision
If multiple analysis regions are used to monitor different cable coil portions, then the precision of winding error detection improves, but the processing time and computational load increase
Solution Approach 1:
The system analyzes only specific partial regions of the cable drum image rather than processing the entire image uniformly. By focusing on radially outer regions for gap detection, radial edge regions for skip detection, and axial end regions for stacking detection, the system achieves high precision for specific error types while reducing overall computational burden compared to full-image analysis.
Data Source
AI summary
The disclosure relates to a device for monitoring the winding behaviour of a cable, comprising a cable winch with a rotatable cable drum on which the cable is mounted so that it can be wound and unwound, at least one sensor unit configured to capture a recording of two-dimensional images of cable coils on the cable drum, and an analysis unit configured to receive the recording from the at least one sensor unit, define at least two different analysis regions in the recording, each representing at least one cable coil in whole or in part, and analyse these independently of one another for the presence of a faulty cable winding. The disclosure also relates to a working machine, such as a crane, civil engineering machine, or cable excavator, having the device, a method for monitoring the winding behaviour of the cable, and a corresponding computer program product.