2:10pm - 2:35pm Case Study: Real-Time Detection of Small Hazardous Liquid Pipeline Leaks Using Remote Sensing and Machine Learning
Tuesday Jun 26
This presentation discusses the recent research focused on the development of the Smart LEak Detection System (SLED). This system leverages optical sensors and machine learning techniques to reliably detect small hazardous liquid leaks including crude oil, gasoline, diesel and mineral oil, including being able to classify these different substances in real-time on different surfaces and in several different environmental conditions. This session will also cover current work, funded by the U.S. Department of Energy (DOE) National Energy Technology Laboratory (NETL), which is focused on the development of the Smart Methane Leak Detection (SLED/M) system, which can be used to monitor facilities, such as compressor stations, but is also suitable for airborne applications (manned and unmanned aircraft) and other ground applications. SLED/M uses an optical camera and machine learning techniques to autonomously and reliably detect “fingerprints” of fugitive methane emissions with an emphasis on: (1) autonomy (no need for a human to be in the loop), (2) high reliability (low false alarm rates), and (3) real-time performance.