Generator Inspection Robot Image Analysis for Worker Workload Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The workload of inspection workers is high when using inspection robots for generator inspections, as they need to visually check moving images for extended periods, which can be tiring and inefficient.
Innovation Solution
An inspection support device that acquires image data from a camera-equipped inspection robot, uses trained classification and object detection models to analyze the data, and presents only critical images and results to the worker, reducing the need for prolonged visual inspection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If an inspection robot is used to travel inside the generator and capture images, then the generator can be inspected without pulling out the rotor, but the inspection worker still needs to visually check moving images for extended periods, resulting in high workload
Solution Approach 1:
The inspection support device automatically analyzes image data captured by the inspection robot using trained classification models, performing the inspection task itself rather than requiring the worker to manually examine all images. The system serves itself by autonomously determining anomaly classes and selecting critical images for worker review.
Solution Approach 2:
The classification model performs preliminary analysis of image data before presenting it to the inspection worker. By pre-processing and filtering images based on anomaly detection, the system prepares only the most relevant images for human review, reducing the worker's burden of examining all captured images.
2Reliability
If all image data captured by the inspection robot is presented to the inspection worker for review, then comprehensive inspection can be achieved, but the inspection worker faces high workload and extended inspection time
Solution Approach 1:
The system extracts only the critical images containing anomalies from the complete set of captured images using the classification model. Instead of presenting all images to the worker, the system separates and presents only those images that require human review, thereby reducing inspection time while maintaining reliability.
Solution Approach 2:
Different processing approaches are applied to different images based on their content. Normal images are automatically classified and set aside, while images containing anomalies are selected for worker review. This localized differentiation ensures comprehensive inspection of critical areas while reducing overall processing time.
3Measurement precision
If the inspection robot captures and transmits all image data to the inspection device, then complete inspection coverage is achieved, but the data processing burden and inspection time increase significantly
Solution Approach 1:
The classification model performs preliminary classification of image data before it reaches the inspection worker. This pre-processing step automatically identifies and flags images containing anomalies, reducing the complexity of manual review by the worker while maintaining inspection accuracy.
Solution Approach 2:
The inspection support device acts as an intermediary between the inspection robot and the inspection worker. It receives all image data from the robot, processes it through the classification model, and presents only relevant results to the worker, thereby reducing data processing complexity at the human interface while maintaining complete inspection coverage.
Data Source
AI summary
An inspection support device includes: a transmission/reception unit acquiring image data captured by a camera installed on an inspection robot capable of traveling inside a generator; a classification unit using a classification model, which is a trained model for determining class indicating degree of anomaly from the image data, and the image data to determine the class corresponding to the image data; an object detection unit using an object detection model, which is a trained model for identifying foreign substance in the generator from image data, and the image data to determine which object is the foreign substance in an image; a result integration unit using results of determination by the classification unit and the object detection unit to select, from the image data, image data to be checked; and a result presentation unit presenting the image data to be checked and the results of determination to the inspection worker.