Marathon Oil Reduces Intelligent Alert Creation Time from Months to Hours Using AWS Partner Seeq
Marathon Oil reduced the time it took to create a new intelligent alert for its wells from months to hours using AWS Partner Seeq. Marathon Oil, an independent exploration and production company specializing in oil and gas, wanted to improve scalability and reduce time to value for its intelligent alert system. By running its jobs using Amazon EKS and Amazon MSK, Marathon Oil pulls data from Seeq, which uses Amazon EC2 to run its applications and then initiates field actions to keep wells online and limit deferred production. With a goal of improving production performance, Marathon Oil now has a solution that allows its employees to create intelligent alerts for exception-based surveillance without involving the IT team.
Accelerating Time to Value for Marathon Oil Using AWS
As an independent oil and gas production company, Marathon Oil wanted to build intelligent alerts from its time-series production data to deliver actionable insights for its digital oil field application. But because it took months to create an actionable alert, Marathon Oil sought a more efficient solution. Nurturing a culture of innovation, the company was eager to move toward the future with its alert system by embracing the cloud. Marathon Oil’s main goal was to empower its business users to create their own intelligent alerts and provide a similar user experience and process shared across all its assets.
The company began using Amazon Web Services (AWS) in 2019 as its preferred cloud provider for application development and considered building a custom solution to fit this need but wanted to investigate commercially available options first. AWS introduced Marathon Oil to AWS Partner Seeq. After reviewing other solutions and comparing them with Seeq, Marathon Oil quickly determined that Seeq was the only provider to fit its requirements and that Marathon Oil would realize significant cost savings by avoiding custom development. “Seeq has the flexibility and ability to scale and create alerts on the nearly 4,000 wells that we have,” says Mark Betts, IT manager of digital solutions at Marathon Oil. “Seeq was the only solution that could achieve that.”
Choosing Seeq to Improve Intelligent Alerts and Monitoring Systems
The first priority for Marathon Oil was to reduce time to value for creating intelligent alerts, which would warn the company of changing conditions across its wells and facilities. A second priority was to build a common integration hub to integrate its field service management tool to initiate tasks and provide a shared user experience across its different assets. The company’s previous solution took months to develop the code for an alert and implement it, and Marathon Oil wanted more value from deep dive time-series analytics than the existing solution provided. In April 2021, Marathon Oil began to work with Seeq on a pilot solution that was integrated in 3 months. When this gained traction with Marathon Oil’s operations teams, the company signed a deal with Seeq, being Seeq’s first-ever subscription in AWS Marketplace, the place to find, test, buy, and deploy software that runs on AWS. Marathon Oil has been building and expanding on the pilot ever since.
Seeq, founded in 2013, is an advanced analytics application for process manufacturing companies and is specifically designed to work with time-series data. The company’s mission is to help process industries that are data rich but information poor by providing a sole source of analytics to increase time to value and time to insight. Seeq provides a browser-based application running on Amazon Elastic Compute Cloud (Amazon EC2), which provides secure and resizable compute capacity to support virtually any workload. The solution from Seeq fits alongside Marathon Oil’s monitoring solution, and with it, Marathon Oil can monitor assets at scale and perform root-cause analytics.
Improving Scalability and Analytics Using Seeq and AWS
The digital oil field application designed by Marathon Oil collects production data from its nearly 4,000 wells and presents this data in dashboards. Marathon Oil connects this data to Seeq applications to create intelligent alerts. The application that assists this data connection is Seeq Cortex, which powers the applications to provide calculations at scale, data connectivity, and administration features. Seeq Cortex then presents the digital visualization layer of the solution. The analytics are run through Seeq Workbench, which includes features to expedite the full arc of the analytics process. And using its engineering tool, Seeq Data Lab (an application for data scientists and process engineers to access Python libraries to expand their analytics), Seeq has built customized user interfaces for alert management.
Marathon Oil is using Seeq in two ways: scaling alerts to all its wells and driving integration with an internal hub that initiates notifications and tasks in its field service management tool. Marathon Oil runs jobs using Amazon Elastic Kubernetes Service(Amazon EKS), a managed Kubernetes service to run Kubernetes in AWS and on-premises data centers, which the company uses to automatically manage containers in the AWS. To orchestrate the entire process on Seeq Cortex, the company uses Amazon Managed Streaming for Apache Kafka (Amazon MSK), which businesses use to securely stream data with a fully managed, highly available Apache Kafka service. The Amazon EKS cluster runs custom Python jobs to call the Seeq Data Lab API and generate analytics on Seeq Workbench. Then, all the data is pulled back onsite using Amazon MSK before the downstream integration with the oil field service management tool and email notifications. On Seeq’s end, everything runs on Amazon EC2. “This solution is increasingly important every day,” says Betts. “Keeping wells online and limiting deferred production are the biggest opportunities to assist the assets going forward.”
Marathon Oil has significantly improved its time to value for intelligent alerts. The process of creation and integration of a new alert has gone from months to hours. “Now that we’re fully integrated, a user can create an alert in Seeq, access our integration hub to schedule the alert, initiate tasks, and send email notifications within hours,” says Betts. Now, when someone identifies an issue in the field, they can use Seeq to investigate the time-series data surrounding the event, set up an alert to warn of potential issues in the future, and assign a field operator to investigate it, all in near real time.
Using Seeq improves scalability for Marathon Oil by connecting production data from across all its wells. The company has over 50 employees using the solution with 170 workbenches in Seeq Workbench. It generates 1,500 tasks and over 1,500 notifications a month. What was being manually identified in the past is now automatically generated. Overall, by using Seeq, Marathon Oil is looking to increase production by proactively identifying issues to increase uptime.
Approaching Future Innovations alongside Seeq and AWS
By using Seeq, Marathon Oil has improved its time to value and scalability, opening new opportunities for the company. Marathon Oil is looking to push its number of alerts from the dozens it has into the hundreds. Being able to scale with Seeq and AWS is an important focus for the future of the company. Marathon Oil wants to enhance the overall reporting and visualizations for dashboards, both within Seeq and without, so that its employees can drill down from a macro level for all its wells. “We are pushing the limits from a performance perspective,” says Betts. “What we’re most looking forward to is Seeq working on its horizontal scaling, using AWS to scale out and support multiple workloads and multiple processes. We will be one of the first customers using this solution.”
About the AWS Partner Seeq
Seeq is an advanced analytics application for process manufacturing companies specifically designed to work with time-series data. The company’s mission is to help process industries that are data rich but information poor by providing a sole source of data for manufacturers to improve time to value