Hello from Alaska!
Last week industry leaders, experts, and innovators gathered at the Houston Aquarium for the OpenText Energy Summit in Houston, a pivotal event driving conversations on the intersection of artificial intelligence, information management, and energy sector dynamics.
The summit provided valuable insights into the future of energy operations with AI.
Let’s dive into the key takeaways from this event:
Physics-based models vs. data-driven models: Insights from Arun Narayanan, CIO of RES Group
Arun Narayanan, the distinguished keynote speaker and Chief Digital Officer at RES, detailed the renewable energy strategy he spearheads within his organization. His presentation delved into the comprehensive approach of RES in executing wind, solar, and storage projects on a global scale. Arun emphasized the core pillars of this strategy, center on maximizing performance, streamlining operational efficiency, and elevating safety standards through innovative solutions, thereby unleashing the full capabilities of renewable energy initiatives.
Arun highlighted that physics-based modeling has stood the test of time and that when your business problem and dataset is grounded in physics, physics-based models will deliver the answers you’re looking for. When your business problem is not grounded in physics, data-driven AI models are tools that can help solve those problems.
Arun emphasized to be sure to have a stated business problem and the business outcome that is expected by solving the problem.
Great AI needs great information management
Phil Schwarz, OpenText Industry Strategist for the Energy sector kicked off the next section, Artificial Intelligence for Energy: To ensure safe and reliable energy, great AI requires great information management.
There was a good mix of customers with IT and non-IT roles which was fantastic because outstanding information management is done collaboratively between IT and each line of business.
Phil kicked off with four themes that the future of energy consists of. Each theme has direct ties to the information management domain:
- The enterprise is reinvented: Every energy company is in a sense now a software company. Whether it’s barrels/day, mcf/day, btu/day, kilowatt-hours, the speed of energy is predicated on digital speed and that requires information to be organized, integrated, automated, and secured.
- Stewardship is redefined: Energy companies are operating and maintaining 50-100% more assets per employee than a decade ago and this ratio of man and machine will continue to rapidly rise. Organizations will expect themselves and by the marketplace to advance along their journey to zero safety incidents. All HSE incidents are preventable and being a trusted brand in the energy sector will not only have the capability to grow energy incrementally and reliably but to do it accident free as well.
- Human-centric work: In some areas of the energy sector up to 50% of the workforce is expected to retire over the next decade. Gen X/Y/Z are dominant in the workforce and their expectation is for information to be instant and in context for them to succeed in their role.
- AI is mainstream: Critical spare parts, replacement equipment, and field services will automatically be ordered, and its status and location can be tracked at any point along the supply chain. Predictive maintenance will be enhanced and gain ground on other maintenance methodologies. Hazard identification from pipeline leaks, vegetation overgrowth, location site security, and many other use cases will automatically get detected through image analytics powered by AI and machine learning.
The exponential advantage of AI in energy – use cases
Phil shared three success stories with traditional (non-generative) use cases, followed by a handful of generative-AI use cases.
Use case #1
First, Phil shared the success story of Western Midstream. Using OpenText™ AI & Analytics solutions, they analyzed and classified large volumes of paper and electronic documents, identifying key attributes and search terms and determining the business value of the content. Not only did they automatically enrich their operational content using AI but also migrated their content to an industry leading enterprise content management platform. By leveraging AI, they were able to quickly create a single source of trusted information, unlocked major time savings and productivity, and simplified compliance with stringent regulations.
He expanded while this was an M&A type business event, this use case is scalable for brownfield projects where asset documentation may have lacked governance for decades and this could be a good initial step to organize information pre-FEED (front-end engineering design).
Also, this is a use case that can be used to scale across business units. Perhaps one business unit has modernized their content but another has not. AI is a great tool to get business units lagging in information management best practices organized and modernized quickly.
Use case #2
Next, Phil shared a story of his father-in-law working as a pipeline & civil maintenance supervisor on the pipeline that transports crude oil from Prudhoe Bay down to Valdez, Alaska.
His father-in-law spent 25 years on the pipeline and part of his role was to get in a helicopter once per week to do an aerial inspection of the 100-mile section of the pipeline he was responsible for. Similar inspections were done for the remaining sections of the 800-mile pipeline. This story teed up how combining the use of AI with unique ways of obtaining imagery through drones, satellite, and other means, hazard identification can be done across more assets, at less cost, and with higher hazard detection rates. This story went on to transition into a second success story shared.
A global integrated energy company was having troubles with uncontrolled oil theft from their pipelines in Africa. Not only did this create a significant operational risk, but it was also a safety risk for those involved with human surveillance of the pipeline. This super major used OpenText AI on satellite imagery to detect oil bunkering. Oil bunkering in this region is a location where thieves dig underneath the pipeline creating a pile of dirt adjacent to the pipeline. And then they perform a ‘hot-tap’ procedure to drill a hole and install a simple valve into an active operating pipeline.
The implementation of AI in this use case involves processing high-quality satellite images devoid of oil bunkering through OpenText software. Through this process, the system rapidly learns to recognize and identify the characteristics indicative of a well-maintained pipeline condition. Afterward, satellite images depicting instances of oil bunkering are analyzed, and utilizing known object classification techniques, the customer can pinpoint these locations. Subsequently, they collaborate with local authorities to take appropriate action against the oil theft from the pipeline.
This use case is scalable to vegetation management for utilities, visual corrosion monitoring, site security for automatically tracking license plates, and others across the energy sector.
Use case #3
The third AI success story shared was of the Chamber of Electric Energy Commercialization (CEEC). The CEEC is the Brazilian electricity operator in which two-thirds of the country’s electricity come from hydro. Their electricity pricing model was done a weekly basis and they were looking to move to an hourly Time-of-Use (TOU) pricing model to align supply costs with consumption. In their previous model, pricing queries could take more than five hours and that was just not doable in an hourly pricing model.
Using OpenText’s AI advanced analytics solution, prices are closer to the operating reality as they are calculated daily, on an hourly basis, and published the day before the operation. Their price curve tends to follow the real demand curve more realistically, reducing the need for additional charges and significantly minimizing the overall costs of the system.
Turning to OpenText™ Aviator
Phil covered six examples of OpenText Aviator, a family of generative-AI capabilities that leverage large language models (LLMs) and private data sets to solve cases. He provided three examples of how he would have used this technology as a former field engineer in the energy sector:
- Find asset documentation faster with an AI-powered assistant: “A lot of times I didn’t need a document per se, I needed information within a document such as equipment specifications, operational steps, safety information, and more. Through chat-based, conversational search, I can quickly find the answers to those questions and the documents that contain those answers.”
- Enabling users to easily search and access supply chain information: “Prior to going offshore I spent a significant amount of time tracking down the location of spare parts, the status of equipment that was in route, and other supply chain related information. Generative-AI can be used establish secure connectivity with your suppliers so that you don’t need to pick up a phone or send an email to have visibility into the status of spare parts, replacement equipment, and service personnel to keep critical energy assets running.”
- Redefining Tier 1 support with a generative AI as subject matter expert: “Many departments have some type of tiered support system to get answers to questions fast. One of the best examples I’ve encountered in my career is an application developed by Schlumberger called InTouch and it was designed to connect the field with experts so that if I were on an operation and needed an answer to a question fast, there was an avenue to achieve that. What happened more than 90% of the time is that the question I had already had an answer logged within InTouch for another field engineer halfway around the world. Instead of scanning tier one support tickets to find answers to questions, generative-AI can find answers to those questions from those logged cases.”
“The Last Frontier” Analogy: Alaska vs. AI
On an inspirational note, Phil shared analogy between his home state of Alaska and artificial intelligence.
Alaska’s land mass encompasses over 570,000 square miles with a population density of just over one person per square mile. Alaska is often called the last frontier because of its expansive areas of land that still have not been accurately chartered, mapped or explored.
The same can be said for information but unlike Alaska, information is vastly growing at a very fast pace. In OpenText CEO & CTO Mark Barrenechea’s book, Versant, he states that “machines now generate one million times more information in one day than all humans on this planet do in an entire year.” All industries have machines however there is no industry as machine or capital intensive as the energy sector.
Like Alaska, AI presents endless possibilities and uncharted territories for innovation and discovery in all industries and especially so across the energy sector. Like the pioneers who ventured into Alaska’s wilderness, more and more of your people will be venturing into AI. Information Management is a journey and great AI will requirement great information management. On behalf of all of OpenText, we’d welcome the opportunity to be your strategic partner along your journey.
Huge thank you to all of the attendees and OpenText’s solution extension partner, Fastman, who sponsored the event.
