Tubular Location Tracking via Machine Learning Vision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Robotic systems in subterranean operations face challenges in efficiently managing and tracking tubulars due to the lack of accurate location information, leading to inefficiencies and potential safety risks during handling and manipulation on the rig.
Innovation Solution
A system utilizing a combination of imaging sensors, machine learning modules, and rig controllers to determine the estimated location of tubulars by processing images and calculating deviation parameters, allowing for precise tracking and management of tubulars during subterranean operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If robotic systems operate without accurate location information of tubulars, then the system complexity is reduced, but the productivity and operational efficiency deteriorate due to hunting behavior
Solution Approach 1:
The patent introduces imaging sensors (cameras) as intermediary devices that capture visual information about tubular locations. These sensors act as mediators between the robotic system and the tubulars, providing location data without requiring complex direct sensing infrastructure throughout the workspace.
Solution Approach 2:
The patent replaces mechanical hunting behavior with vision-based guidance. Instead of relying on mechanical search patterns and physical detection, the system uses image processing and computer vision algorithms to determine tubular locations, substituting mechanical exploration with optical sensing and computational analysis.
2Measurement precision
If robotic systems implement comprehensive tracking of tubulars, then the measurement precision of location is improved, but the device complexity increases due to additional sensors and processing requirements
Solution Approach 1:
The patent creates visual copies (images) of the physical tubulars and their locations. By capturing images with sensors and processing these visual representations, the system obtains location information without physically interacting with or adding complex tracking hardware to the tubulars themselves.
Solution Approach 2:
The patent substitutes mechanical tracking systems with vision-based detection. Instead of using RFID tags, barcodes, or mechanical markers on tubulars, the system uses natural visual features and image processing to identify and locate tubulars, reducing the need for additional physical tracking components.
3Device complexity
If robotic systems perform manual searching for tubulars, then the device complexity remains low, but the loss of time increases due to hunting behavior
Solution Approach 1:
The patent implements preliminary action by continuously capturing images and pre-processing visual data to identify tubular locations before they are needed. The system proactively builds and maintains location information through image processing, so when a robotic operation requires a tubular, its location is already known, eliminating search time.
Solution Approach 2:
The patent replaces mechanical searching and hunting behavior with vision-based location determination. Instead of physical exploration and trial-and-error detection, the system uses image capture and computational analysis to instantly identify tubular positions, dramatically reducing the time required to locate equipment.
Data Source
AI summary
A method for conducting subterranean operations can include engaging a tubular with a pipe handler, moving the tubular with the pipe handler to a new location, disengaging from the tubular at the new location, determining, via a rig controller, an estimated location of the tubular based on the new location at which the pipe handler disengaged from the tubular, determining, via a machine learning module of the rig controller and one or more imaging sensors, a deviation from the estimated location of the tubular.