AI Pipe Mapping With INS Data for Underground Asset Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems lack the ability to automate the process of collecting and accurately predicting characteristics of buried or hidden pipes using large data sets, and existing technologies are slow and inefficient in processing and analyzing such data.
Innovation Solution
Utilizing a combination of imaging and inertial navigation system (INS) sensor data with voice annotation, processed by a neural network using deep learning, to create a training database for predicting pipe characteristics and generating a 3D map of underground utilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional data collection and analysis methods are used for pipe inspection, then the process is simple and easy to implement, but the speed of data processing and analysis is slow and efficiency is low
Solution Approach 1:
The patent replaces traditional mechanical data processing methods with artificial intelligence and deep learning algorithms. The neural network automatically analyzes pipe inspection data, replacing manual or rule-based processing systems, thereby dramatically increasing processing speed while managing complexity through automated intelligent systems.
Solution Approach 2:
The system employs self-learning neural networks that automatically improve their analysis capabilities through continuous training on pipe inspection data. The AI system serves itself by automatically identifying patterns, features, and anomalies without requiring constant human intervention or manual programming of analysis rules.
2Measurement precision
If a single data collection method is used, then the system is simple, but the accuracy and completeness of pipe characteristic prediction is limited
Solution Approach 1:
The patent merges multiple data collection methods including imaging data, inertial navigation system data, and voice annotation data into a unified analysis system. The neural network processes these diverse data sources together, allowing the system to achieve high prediction accuracy by combining complementary information from different sensors and input methods.
3Loss of time
If manual data annotation and analysis is performed, then the system requires minimal technology, but the time required for data processing is excessive
Solution Approach 1:
The patent replaces manual data annotation and analysis with automated artificial intelligence systems. The neural network automatically processes inspection data, identifies pipe characteristics, and generates predictions without human intervention, reducing analysis time from potentially days or weeks to minutes or seconds while implementing high-level automation.
Solution Approach 2:
The system performs preliminary data processing and feature extraction automatically before final analysis. The neural network pre-processes incoming data streams, identifying relevant features and patterns in advance, which accelerates the overall analysis process and reduces the time required for comprehensive pipe characteristic assessment.
Data Source
AI summary
Systems and methods are provided for recognizing and mapping features inside of pipes using Artificial Intelligence (AI). In an exemplary embodiment, a pipe inspection camera system including Inertial Navigation Systems (INS) and other sensor capabilities is inserted into a pipe to collect data that can be provided to a Deep Learning model to build a training set. Training data may include newly collected data and/or historical data. The model may be trained based on collected sets of training data, testing data, and/or user predefined classifiers. The Deep Learning model may use thresholds to determine if a set of data falls within a specific class. Classes may be based on data related to pipe features such as size, shape, material, age, routing or connection features including bends and/or joints, etc. AI data may be processed locally in the pipe inspection camera system, remotely on a mobile device, and/or in the Cloud.