Sept. 20, 2021
Developing faster and more accurate cross bore detection
A group of Spartan Engineers had an impactful research field test recently that included an advancement from the MSU Nondestructive Evaluation Laboratory (NDEL).
Their work centered on the detection of cross bores -- a widely-existed high-level threat for the safety of the gas industry and the general public where gas pipelines were installed unintentionally through the sewer pipelines. Without discovery, they can cause sewer blockages leading to gas explosions, because the gas pipe may be cut through by a drain auger cleaner when cleaning out the sewer pipe blockage.
The current state-of-the-art tool capable of this task, a sewer camera, is quite limited due to the various blockages and substances in the sewer lines, said Yiming Deng, an associate professor in the NDEL, which is part of MSU’s Department of Electrical and Computer Engineering (ECE).
Because of these limitations, Deng said there is an urgent need for a reliable cross bore detection tool in challenging field environments.
After receiving a grant from the Gas Technology Institute (GTI), Deng and his team developed an artificial intelligence (AI) based multi-channel electromagnetic (EM) sensing system for rapid and high accuracy cross bores detection, nondestructively.
Postdoctoral Research Associate Guanyu Piao and Jiaoyang Li have been working closely with Deng on the project’s model study, sensing system development and AI-based classification algorithm.
“The multi-channel EM sensing system is designed to move through the gas pipe to identify and locate cross bores buried in different soil types with high sensitivity,” Piao explained.
“An advanced machine learning algorithm was developed to improve the probability of cross bore detection and the classification accuracy of surrounding material types,” Li added.
The team tested this novel multi-channel EM sensing system in the field at GTI in Chicago earlier this year. Deng said the EM sensing system successfully identified technical gaps -- cross bore, butt fusion, saddle fitting, clamp and other issues -- with high sensitivity and robustness in real-time.
Adithya Rao, a senior in computer engineering, was able to travel with the research group for the field test as part of his EnSURE summer program experience.
“During our visit, I assisted in setting up the demonstration and collecting data in a real-world scenario which was emulated at the company’s headquarters,” Rao said. “I learned about the importance of considering the various real-world factors that may be missed when working in a controlled lab environment. The experience as a whole was amazing, and I am so grateful to have been a part of it.”
Deng called the field test results “very exciting” and the novel electromagnetic sensing system developed by MSU a “pioneering and promising technology to solve the challenging cross-bores identification problem. We will continue to optimize and advance the sensing system and commercialize this wonderful technology to lead the gas industry,” he added.
Varun Desai, the project lead for GTI, said the work and collaboration with MSU advanced a critical technology addressing a high-risk gas industry problem.
“The team was receptive to the industry needs and designed effective prototypes to detect cross-bores. Despite the challenges due to pandemic related disruption, Dr. Deng and his team were agile enough to finish the project on time,” Desai said.
“The field test day was a success, the prototypes were extensively tested, and they worked well, providing a field proof-of-concept,” Desai added. “During the visit, MSU postdocs and students were eager and curious to know the work GTI does to make the gas industry safe and efficient. I hope they had a good time learning and testing, despite the rain on that day.”
Written by Kee-Ri Burkitt, student writer in the College of Engineering