PRCI/PNNL WEBINAR (RECORDING): Machine Learning-Based Prediction of Future Pipe Condition

Tue, October 13, 2020

Click here to view the webinar recording. Webinar runtime: 73 minutes
Join PRCI and the Pacific Northwest National Laboratory as we share a New Research Project: Predicting Future Pipeline Condition with a Model built on Past Inspection Data, Operating History & Operator Experience using Machine Learning.  PNNL is actively working on this project an looking for data to assist in their research.
Presenter: Kayte Denslow, Pacific Northwest National Laboratory
Moderator: Gary Choquette, PRCI
This webinar will summarize a new project – at its start – that aims to combine decades’ worth of existing pipeline inspection/operating history data (and experience) with the machine learning and computing resources at PNNL to build a “data-driven/physics-based” model of pipeline corrosion. The model, which will be developed with the help of industry partners, would be made available to pipeline operators/ILI companies to help predict the future condition of pipelines and support decisions on where and when to perform preventive maintenance to mitigate risks of pipeline failure.
The envisioned model will be described during the first half of the webinar and be followed by 1) an invitation to partner on the project, 2) a description of benefits to data partners, and 3) past examples of successful PNNL-industry partnerships in data sharing. The project is being funded by the Office of Fossil Energy at DOE and led by PNNL.
Key Learning Points include:
Description of the project and the envisioned model
Benefits to project partners (operators, ILI companies)
Examples of successful industry data sharing 
Attendance is limited to the first 500 registrants to join the webinar. All remaining registrants will receive a link to view the webinar recording.
After registering, you will receive a confirmation email containing information about joining the webinar.