logo
menu

Digging data not ditches

The Trans-Alaska Pipeline crossing a ridge, Tanana River is in the foreground (via Wikimedia Commons, by Beeblebrox).
The Trans-Alaska Pipeline crossing a ridge, Tanana River is in the foreground (via Wikimedia Commons, by Beeblebrox).

A OneBridge-Phillips 66 joint development project is set to advance digital management and machine learning for pipelines into the cloud.

The US Pipeline and Hazardous Materials Administration reports that since 1997, in Natural Gas pipelines alone incidents have resulted in 322 deaths and damages totalling to over $7 billion. Ageing infrastructure and increased public scrutiny on how energy companies conduct their business means that being able to know what a pipeline looks like and where the problems are is only becoming more important.

In September 2017 OneBridge, a software company out of Alberta, Canada, supplied the Houston-based diversified company, Phillips 66 with something to address this issue: a machine learning and data science system called the cognitive integrity management (CIM) solution. Now both companies are looking to build on the system by moving the software into an online app, dubbing it the ‘integrity management solution’ (IMP). The IMP is now in development, and is aimed at catering to the needs of major pipeline operators.

The previous pipeline management solution was a combination of the CIM and Phillips 66’s pipeline data management system (PT-DMS). Phillips 66’s system was developed to be a comprehensive solution to manage its pipelines, combining functions including assessment team scheduling, analyses of data integrity and anomaly worklist tracking.

The CIM built on this by providing data normalisation and alignment, and applying machine learning technology to identify threats to the pipeline. In a statement, the software company said that they reduced the time it took to analyse the data from ten weeks, to two hours. CIM also allowed all of the data to be assessed whereas in the previous manual process could only process 5% of data collected.

OneBridge has patented this machine learning technique.

The resulting boost in awareness of pipeline condition allows companies to streamline their maintenance programme and allow engineers to spend more time on more substantial tasks like risk analysis and corrosion modelling. OneBidge is also preparing to use Microsoft’s HoloLens to enhance pipeline assessment.

In a press release, OneBridge President Tim Edward said: “This development project with Phillips 66 represents an important milestone for OneBridge… Our vision at the outset was to develop a cloud application that enables pipeline operators to manage their pipeline assets as smart infrastructure.”

CTO Brandon Taylor added: “PT-DMS is one of the most comprehensive and sophisticated pipeline management solutions within the industry today, which will ease migration to the cloud and reduce time-to-market for IMP.”

The company expects IMP to launch market-wide in Q4 of 2018.

The Trans-Alaska Pipeline crossing a ridge, Tanana River is in the foreground (via Wikimedia Commons, by Beeblebrox).