Since the end of 2022, US-based Marathon Pipe Line has been using and testing Flyscan Systems’ remote sensing technology over their liquid pipeline Rights-of-Way (ROWs) to benchmark and validate the use of machine learning (ML) algorithms, photogrammetry and hyperspectral imaging to perform real-time threat detection, liquid leak detection of crude oil and refined products, and other advanced imaging features related to geo-hazards.

This paper will present scientific approaches used, and real-life results from 198 patrols between November 2022 and August 2023 which identified 2,594 threats across 35,706 miles of ROW from Alaska to Texas, including the Rockies and the Mid-west.

Examples will include, what is potentially the first ever detection of a diesel seeper leak, identified below computational pipeline monitoring system detection thresholds using a hyperspectral imaging system during routine aerial patrols. Other examples will include detection of contaminated soils from third party activity, detection of real encroachments and documentation of various types of objects and activities on the ROW of interest to pipeline operators. Finally, examples of new imaging capabilities will be presented, including detection of exposed pipes after serious weather events, vegetation analysis, and erosion monitoring.