Submersible ROV Vision-Based Modeling for Transformer Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing submersible inspection systems face limitations in effectively inspecting submerged objects, particularly electrical transformers, due to challenges in data collection and modeling within sealed and contaminated environments.
Innovation Solution
A submersible remotely operated vehicle (ROV) equipped with cameras, motors, and wireless communication, capable of wireless remote control and vision-based modeling, allows for in-situ inspection and data collection within transformers, using cameras to capture images, motors for navigation, and wireless communication for data transmission, while employing vision-based modeling systems to generate accurate 3D models of transformer components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If traditional inspection methods are used in sealed and contaminated environments, then accessibility is limited, but inspection accuracy and data quality deteriorate
Solution Approach 1:
The patent replaces traditional mechanical inspection systems with a submersible ROV equipped with vision-based sensing systems. The ROV uses cameras and optical sensors to capture images and construct 3D models of transformer components, eliminating the need for physical contact or direct human access to contaminated environments while maintaining high inspection accuracy through digital modeling and image processing techniques.
2Measurement precision
If submersible ROV with vision-based modeling is deployed, then inspection accuracy improves, but system complexity increases
Solution Approach 1:
The submersible ROV is designed as a multi-functional platform that integrates navigation, imaging, data processing, and 3D modeling capabilities into a single system. The vehicle can perform multiple inspection tasks including visual inspection, dimensional measurement, and defect detection using the same hardware platform, thereby managing complexity through functional integration rather than separate specialized devices.
Solution Approach 2:
The vision-based modeling system creates digital 3D copies and virtual models of transformer components instead of requiring physical access or manipulation. These digital replicas allow for accurate inspection, measurement, and analysis without the complexity of physical intervention systems, transferring the inspection function to the digital domain where complex environments pose no barrier.
3Productivity
If real-time data processing is implemented, then inspection efficiency improves, but computational requirements and energy consumption increase
Solution Approach 1:
The data processing workflow is segmented into multiple stages: on-board preprocessing of images and videos during the inspection mission, selective transmission of key data to ground stations for intensive processing, and post-mission comprehensive analysis. This segmentation allows real-time processing of essential data to maintain inspection efficiency while avoiding continuous high-energy computation that would deplete the ROV's power supply.
Solution Approach 2:
The system performs partial real-time processing by prioritizing critical inspection data for immediate analysis while deferring less urgent processing tasks. The vision-based modeling system continuously updates 3D models but focuses computational resources on detecting and analyzing defects rather than processing every pixel in real-time, achieving sufficient inspection efficiency with reduced energy consumption.
Data Source
AI summary
A submersible vehicle which includes a plurality of cameras can be used to collect visual images of an object of interest submerged in a liquid environment, such as in a tank (e.g. transformer tank). In one form the submersible vehicle is remotely operated such as an ROV or an autonomous vehicle. Image information from the submersible along with inertial measurements in some embodiments is used with a vision based modelling system to form a model of an internal object of interest in the tank. The vision based modelling system can include a number of processes to form the model such as but not limited to tracking, sparse and dense reconstruction, model generation, and rectification.