AI and Machine Learning

With AI, zero failure is more than a pipe dream

Improve confidence with a complete view of data.
By Tim Edwards and Rob Salkowitz February 4, 2019
Figure 1. Artificial intelligence is being applied to pipeline management based on its ability to identify patterns in data that are not always intuitively obvious to humans. Courtesy: OneBridge

Everyone in our business knows—or ought to know—about the pipeline maintenance crisis that puts billions of dollars, lives, property, and the reputation of midstream oil & gas industry at risk, leading some in the public to call it a “ticking timebomb.” Statistics indicate tens of thousands of miles of pipes decades beyond their predicted end-of-life, scattered so wide and buried so deep that just finding them on a map can be a problem.

No one is happy with this situation, but it’s not easy to solve. Pipeline integrity teams already are asked to perform miracles with the data generated from traditional inline inspection (ILI) tools, to analyze vast spreadsheets that typically represent only 5% of the data collected. What if the answers are in the other 95%? What happens when new laser scanning technology increases the volume of data exponentially, without any new tools to make sense of it all? And what happens when the senior-level experts who have been keeping everything running for decades retire to spend their days playing golf and passing time with their grandchildren?

Those are the kind of problems keeping risk managers, CFOs, and CEOs awake at night. Today’s expertise and technology is, at its very best, able to hold the line against catastrophic failures. Given the challenges ahead, is it even conceivable to imagine reducing failure risk to zero?

We think it is. Here’s why.

Digital transformation for pipeline operators

Folks use the term “digital transformation” as a buzzword for the investments being made in technologies like artificial intelligence (AI), machine learning, augmented reality, robotics, and wearables taking place across the business world. McKinsey and Accenture recently estimated that digitalization has the potential to create as much as $1 trillion in value for oil & gas companies.

Figure 1. Artificial intelligence is being applied to pipeline management based on its ability to identify patterns in data that are not always intuitively obvious to humans. Courtesy: OneBridge

Figure 1. Artificial intelligence is being applied to pipeline management based on its ability to identify patterns in data that are not always intuitively obvious to humans. Courtesy: OneBridge

AI may be the technology promising the most impact—particularly when applied to pipeline management. AI systems aren’t actually “intelligent,” of course, but they are trained to recognize in Big Data sets patterns that look like problems or issues worthy of human attention. At scale, they can churn through billions of data records to spot everything from suspicious network activity to fraudulent financial transactions, if they know what to consider a problem. The more data these systems look at, the more precise and confident they become in finding what they’re looking for.

That has a bunch of big payoffs for pipeline management. First, you will analyze 100% of your data, as compared to the likely 5% that organizations leverage today to assess safety risk, and can increase that analytical capacity at almost no marginal cost. In fact, costs will probably be reduced by freeing up analyst time and effort by automating the most time-consuming, repetitive parts of their job.

Second, depth of confidence improves based on a more complete view of the data. Think of the millions of dollars that pipeline operators waste every year because they don’t believe their data and err on the side of safety by digging where they don’t need to. Sure, it’s better to be safe than sorry when it comes to risk exposure. On the other hand, there are financial costs to being over-cautious—costs that organizations can eliminate by having clear and precise data on the exact status of every square inch of pipe along thousands or tens of thousands of miles of run.

Bringing data into the business

When we first started OneBridge in 2015, we felt we had a good handle on the technology required to automate the repetitive tasks related to pipeline data analysis that experts hate. Now that we’ve spent several years working with some of the top companies in the business, more capabilities have been added to our system to benefit operators on everything from assessment planning to integrity compliance to threat monitoring. With each new release, including the latest, Cognitive Integrity Management 3.1, the product is getting smarter about how it analyzes and identifies problems, and smarter about how to make adoption easier for customers.

These capabilities take the platform beyond the basic ability to interpret pipeline data by bringing that data into the mainstream of the business. For workgroups and managers responsible for assessment planning, it’s easier to automate, schedule, and monitor many of the project tasks associated with maintenance. Tools have been incorporated that enable regulatory compliance by monitoring a range of technical threats and documenting the integrity of business processes end-to-end.

New monitoring and reporting features have been added that empower teams to collaborate around threat monitoring data, smoothing the path from insight to action. Listening to customers it’s clear they want better reporting, data visualization, and integration with standard analytics tools like Microsoft Power BI. Along these lines, the ability of OneBridge to interoperate with other systems across the enterprise has been enhanced.

Moving forward, part of the mission will be looking for unique and exciting ways to get that data to field crews to make digs faster, more precise, and less costly, so operators can maintain their networks with greater confidence and less expense.

Getting to zero failures

For some people, AI can have sinister connotations, including the replacement of humans with computers. Rest assured, tools like OneBridge are about getting customers to zero failures, not zero employees. For the past 20 years, pipeline integrity experts have been waging a valiant battle against aging infrastructure with inadequate tools and inadequate resources, incurring the wrath of management when they fail, but garnishing little notice or attention in their daily triumphs, despite the odds.

As those people head toward a well-earned retirement, organizations need to embed their knowledge and experience in systems so that their successors can step into new roles without any reduction in operational excellence. Investing in AI-based tools can move that process forward while at the same time it delivers the other benefits of greater confidence, greater data coverage, and reduced risk. It also demonstrates to next-generation workers that the company is willing to invest in modern solutions.

No one can stop the hands of time as they affect our pipelines, organizations, or workforce, but interested parties can get out in front of the problem instead of giving everything they’ve got just to fight it to a standstill. AI is the industry’s secret to stop playing defense and start moving toward the goal of zero failures. It’s not a promise, a vision, or a pipe dream. It’s here today and getting smarter by the minute.

Want this article on your website? Click here to sign up for a free account in ContentStream® and make that happen.

Tim Edwards and Rob Salkowitz
Author Bio: Tim Edwards and Rob Salkowitz, OneBridge Solutions.