Apply to KAUST
Moving to KAUST
Search for Jobs
KAUST Ph.D. students Kai Lu and Yuqing Chen recently won Best Presentation awards from the Society of Exploration Geophysicists. The graduate students are pictured here on the KAUST campus. Photo by Meres J. Weche.
By Tanya Petersen, KAUST News
Artificial intelligence (AI) and machine learning (ML) are now embedded in the global science and technology lexicon, as well as in many industries such as finance, retail, communications and oil and gas. The oil and gas industry, in particular, has recently been digitally transformed, driven by advances in artificial intelligence and machine learning, with applications in areas including seismic processing and interpretation, reservoir characterization, drilling, completion and production forecast.
At a workshop held in Beijing in September 2018 by the Society of Exploration Geophysicists, two KAUST Ph.D. students won Best Presentation awards in separate sessions focused on AI and ML.
Kai Lu and Yuqing Chen both work under the supervision of Gerard Schuster, KAUST professor of Earth science and engineering.
Lu's research centers on machine learning applications in seismic processing—and particularly on developing theoretical techniques on seismic processing and conducting seismic field experiments.
His presentation entitled "Auto-windowed Super-virtual Interferometry via Machine Learning: A Strategy of First-arrival Traveltime" focused on a technique that can significantly improve the efficiency of seismic data processing in the oil industry: connecting super-virtual interferometry. This improves data quality and machine learning methods to automatically enhance data, so that for even noisy seismic data, traveltimes can be auto-picked.
"I feel happy and proud. This award gives me confidence that my research is on the right track and people like my topic and the work I am doing," he stated.
KAUST Professor Gerard Schuster (pictured here) supervises KAUST Ph.D. students Kai Lu and Yuqing Chen and is part of the University's Earth Science and Engineering program. File photo.
"For a large data set, this task becomes extremely time-consuming, which may take people weeks or months to finish," Chen explained. "I use the clustering method, which is an unsupervised machine learning algorithm to automatically pick the semblance spectrum. In this case, we can finish the picking task in hours and also relieve human labor."
"My workstation is quite fast which quickly runs my programs," he continued. "The Shaheen II cluster is even more powerful, which significantly facilitates my work on some really large projects. A supercomputer like Shaheen II allows me to quickly test my programs, which greatly facilitates my research progress. KAUST also has great library [of] E-resources, so I can find and download almost any paper that I'm interested in."
"I'm very happy for these...two hard-charging students who, time after time, come up with their own ideas and follow them until it's either time to give up or pick the fruits of their labors," Schuster said. "In this case, their harvest was bountiful."