UD Software To Identify Oil and Gas Pipeline Threats

Mar 21, 2016

University of Dayton researchers are developing computer software that will automatically detect threats or damage to oil and gas pipelines, thanks in part to the support of one of the nation's largest utilities.

California-based Pacific Gas & Electric has given UD's Vision Lab $254,362 to help researchers continue to create an automated monitoring system for building change detection on pipeline corridors by processing and transmitting images in virtual real time. The system will also report the threat location on a Google map. The Pipeline Research Council International (PRCI) had originally supported the research on machinery threat detection on pipeline right-of-ways. The automated system for threat object detection will also report the threat location on a Google map. "

The project is well on its way. "A vendor is testing the software and results look very promising", according to Vision Lab Director, Vijayan Asari. He says, "We are in the process of finding appropriate customers for commercialization of the product."

Right now utility companies do it the hard way. They have employees watching aircraft surveillance video to detect threats. It’s very tedious work and it’s easy to miss things as graduate student Dan Prince explains.

“If any of these threats go undetected what could potentially happen is the pipelines underground could break and damage the environment and kind of a lot of company resources could be lost. And we’ve seen a few kind of high profile cases like these. This is kind of related to the BP incident, in a different way. We want to protect pipelines that go underground, so that the same type of incident doesn’t happen.”

Students have developed an algorithm specifically for oil and gas pipelines which, in virtual real time (assessing each frame in 1/3 second) processes and transmits images and reports the threat location on a Google Map. Eventually they will have access to satellite imagery.

Credit UD

Vision Lab algorithms have detected 90 to 95 percent of images with a low rate of false positives. Threats are seen as anything that’s unauthorized by a utility company."

Vision Lab scientists have been working on pipeline safety since the lab opened in 2010.