Pipeline Integrity Management and Data Science Blog

Developments with Cognitive Integrity Management

We are excited to share the most recent progress that has been made to our Polaris release set for launch before the end of the year. Our data science and development teams are busy working through user stories for our Polaris minimally viable product (MVP) release. At the end of May we held our kick off meeting with operators participating in our Private Preview at the Microsoft Technology Center in Houston. Since then we have received invaluable feedback and guidance from these operators to ensure we release a product best suited to help the industry optimize business value resulting from the solution.

Success through operator expertise:

The operators participating in the Private Preview are not only helping us, we believe they are shaping what digital transformation means for Pipeline Integrity Management. We cannot express how appreciative we are for their support through this initiative. Further, it’s given us confidence that this release will seamlessly integrate within integrity management processes for operators across the industry. Most importantly, we judge our success as a measure of the value our solutions provide organizations in solving “real” business problems. To that end, in the coming weeks we’re excited to have participating operators start to move existing production business processes into the next version of our Cognitive Integrity solution. We are confident that they will be eliminating a great deal of manual work stored in Excel spreadsheets, making it easier for integrity engineers to focus their expertise on highly valuable analysis – to us, this is the most satisfying outcome.

Future Advancements:

As a result of the quick progress that we have made, we kicked off our next development sprint, Orion, last week. We mapped out the integration points for bringing in our data science and machine learning algorithms to this next version. In addition, as is the case with all SaaS companies, we’ve been continually evolving our existing algorithms. It’s exciting to be able to see our original alignment algorithm evolve to the point that we can now do segment-wise alignment. Operators will be able to define an entire segment and drag and drop every assessment ever done for that segment onto the solution, and have the algorithm align everything in one shot. This evolution will provide a leap forward in Cognitive Integrity Management’s learnings and sophistication with alignment. We are already in the phase of testing this and will provide further updates as the results begin to roll in!

Overview of Cognitive Integrity Management 4.0.


Finally, we’re starting conversations with partners outside of in-line inspection (ILI) data. Although, that data is telling us we haven’t completed leveraging ILI to its fullest extent, we are now incorporating data models into our solution design to integrate other data sets into our learning. As an example, we’ve recently held discussions with ESRI to figure out API’s and other ways to bring in the wealth of data being stored in their systems residing at pipeline operators. It’s sitting there waiting for us to corelate it to ILI data and utilize our data science algorithms to determine if we can increase confidence levels within our probabilistic models around potential threats. We’ll seek out and continue to evaluate where and how additional data can further our mission.

We appreciate everyone that is participating and/or following us on our journey toward the industry’s ultimate mission of predicting pipeline failures to save lives and protect the environment, and we hope to do it with the assistance of machine learning. The metrics we are now gathering from field data generated as a result of analyses performed by operators using our solution are very encouraging.

Please feel free to leave a comment or send us feedback.