A team of five graduate students from the John and Willie Leone Family Department of Energy and Mineral Engineering in Penn State’s College of Earth and Mineral Sciences won first place in the 2024 Chevron National Engineering Competition. The annual competition challenges teams to present novel ideas about contemporary subjects in the petroleum and energy industry, with this year’s topic focused on use cases for implementing artificial intelligence (AI).
Nicolas Bueno, Nijat Gasimli, Ianna Gomez Hanif Yoga and Baran Yucel, whose team took second place the previous year, were part of this year’s winning team.
“It's important to try new things,” said Bueno. “I came to Penn State from Colombia to research potential CO2 and hydrogen storage solutions via studying how bubbles interact in porous media, but I wanted to try something beyond my research and learn about different perspectives.”
“We reviewed the previous topics and brainstormed a lot of potential directions,” said Gasimli, whose research focuses on experimental methodologies for enhanced oil recovery. “But AI wasn’t one of them. It was an exciting surprise.”
The members all agreed that the topic's broad nature would be demanding. AI can be utilized across every department in the oil industry, from operations and reservoir management to finance and human resources. The team knew they would need to be strategic about which aspect they focused on.
“I was very excited because I’ve been working directly with AI the past three years, but I knew the competition would be difficult for different reasons,” said Yucel who utilizes AI in his work to improve geostatistical modeling. “Firstly, the timing is tight and you have to squeeze a lot of complex ideas into a concise presentation. Second, you must balance exploring the different AI applications while maintaining a clear, big-picture perspective.”
The team decided to divide and conquer. Each member focused on the key strengths, weaknesses and potential uses for AI across every industry cycle stage. The decision to tackle the entire lifecycle required each member to become a convincing expert in their respective section and address concerns ranging from job security to accuracy. Their suggestions would also have to be practical, not just theoretical, to appeal to the industrial judges.
Gomez, who has a geophysical and nuclear engineering background, said she relished learning something new. However, she had to balance her research on underground hydrogen storage and reading numerous research papers about geoscience applications.
“The hardest part was to fit everything in a fifteen-minute presentation where we each had to present a section,” said Gomez. “It took a lot of time to filter through everyone’s great ideas to find the right point to highlight. However, after a lot of rehearsals, we managed to make the presentation in only thirteen minutes and leave time for questions."
According to Yoga, refining their presentation was key to their success. The opportunity to receive feedback and direction from their advisers and department experts proved invaluable. Turgay Ertekin, professor emeritus of petroleum and natural gas engineering and former department head, met with the team for several two-hour meetings and helped guide their presentation.
“AI is such a hot topic and it’s talked about as this new thing,” said Yoga. “In reality though, Dr. Ertekin has been researching the use of AI applications since the 90s. We were so thankful to consult with him.”
Ertekin said he was happy to help and meet some of the young, bright minds in the department. In his forty years of experience, he has seen the adoption of many new technologies, and has always encouraged the need to adapt, or add, another tool to the toolbox. To Ertekin, the key is to not abandon, or completely divorce oneself from the tools you already have.
“Petroleum engineers have always found themselves in the middle of ill-defined problems,” said Ertekin. “Imagine trying to solve a 10,000-piece jigsaw puzzle that has 200 pieces missing. Which pieces are missing, what shape do they have, and where do they go — and there’s still the thousands of other pieces you must fit together. AI is a tool that can help identify potential solutions or eliminate possibilities to solve the puzzle quicker, but humans need be there to feed it the correct parameters and ensure it’s accurate as they are the final decision makers.”
The team was especially proud of how they bookended the presentation — with an A.I.-generated image to represent the Nittany Lion sitting at a computer. The image perfectly captured the power of AI and the need for human oversight.
“Baran made this cool image of a lion sitting at a computer looking at graphs and models, but on close inspection, you could spot misspelled words and other issues,” said Mendez. “Of special note, the lion is not even an accurate Nittany Lion, another fault of AI.”
The team’s focus on the big picture, and recommendation of a hybrid approach to implementing AI with human supervision as a safeguard and decision maker, proved to be the winning formula. Buenos felt confident they left a great impression on the judges.
“We felt really good after our presentation because we were able to answer very specific questions from the judges,” said Bueno. “Knowing how to present your idea is a fundamental skill, so I was happy to participate and gain that experience. Working with such a diverse, creative team was a great learning opportunity, too.”
Hamid Emami-Meybodi, chair of the petroleum and natural gas engineering program and director of SERS JIP, said he was pleased when he heard the announcement.
“We are proud of the extraordinary research our graduate students conduct at Penn State. This competition has allowed them to showcase the breadth and depth of their AI knowledge and its implementation in the petroleum industry. “
The annual competition is hosted by the Viterbi School of Engineering at the University of Southern California.