AI Powered Renewable Energy Workshop Speakers

Fireside Chat

Monday, April 17, 2023 ( 8:15 - 9 a.m. )
Bobby Tudor
CEO Artemis Energy Partners
btudor@artemisenergypartners.com

Bobby Tudor is the founder and CEO of Artemis Energy Partners, an Investing and Advisory platform focused on companies involved in the Global Energy Markets.

Mr. Tudor is also a Retired Founder and CEO of Tudor, Pickering, Holt & Co. TPH is a leading Energy Investment Bank formed in 2007 with over 100 employees in five offices in the US, Canada and UK. It is widely recognized for its industry leading position in Energy securities research and Energy investment banking advisory transactions. TPH merged with Perella Weinberg Partners in 2016.

Prior to forming TPH, Mr. Tudor was a Partner at Goldman Sachs and a leader of its worldwide Energy practice. Over his 30-plus year career in Investment Banking, he has worked on many of the defining transactions of the period, across most energy subsectors and geographies.

Mr. Tudor is currently the Chairman of the Houston Energy Transition Initiative, which is a consortium of Houston’s leading energy companies working to shape the region’s Energy Transition Strategy.

Mr. Tudor is on the Board of Directors of Puloli, Inc., an early-stage methane detection and measurement company, as well as New ASEAN Energy, a US based company involved in Petrochemicals business in Asia.

Mr. Tudor is the Past Chair of the Greater Houston Partnership and of the Rice University Board of Trustees. He serves on the Board of Advisors for Rice University’s Baker Institute for Public Policy, the National Petroleum Council, the Jones School of Business at Rice, the Carbon Neutral Coalition, and the National Advisory Board for the Tulane Center for Energy Law. Mr. Tudor also serves on the Board of Directors of the Houston Symphony, Good Reason Houston, and the MD Anderson Cancer Center Board of Visitors.

Mr. Tudor holds a BA in English and Legal Studies from Rice University, and a JD from Tulane Law School. He received the Lifetime Achievement Award for the World Oil Council and the Gold Medal for Lifetime Service to Rice University.

Mr. Tudor lives in Houston with his wife, Phoebe. They have three children, Caroline, Margaret, and Harry. The Tudors are passionate supporters of the Arts, Public Education, Social Services, and Public Parks.

Monday, April 17, 2023

Amy Dittmar ( 8 - 8:15 a.m. )

Amy Dittmar
Howard R. Hughes Provost at Rice University

Detlef Hohl ( 9 - 9:45 a.m. )

Detlef Hohl
Chief Scientist Computation and Data Science at Shell
Detlef.Hohl@shell.com

Detlef Hohl holds a Master's degree in chemistry from Technical University of Munich and a PhD in theoretical physics from Technical University of Aachen (Germany). Before joining Shell in 1997, he was senior scientist at the German National Laboratory Forschungszentrum Jülich.

Detlef started at Shell’s Bellaire Technology Center in seismic imaging research, then moved to Promise seismic inversion R&D, and became R&D team leader for Quantitative Reservoir Management in 2006. From 2010-2017 he was General Manager Computation and Modeling where he led a project portfolio in data analytics, computational engineering and materials science, geoscience and petroleum engineering.

Detlef is adjunct professor at Rice University (Computational and Applied Math) and University of Houston, and visiting scholar at the UK National data science laboratory Alan Turing Institute. He held various temporary and visiting positions at NCSA, SISSA Trieste, NIST and Stanford University.

Daniel Cohan ( 10:15 - 11 a.m. )

Daniel Cohan
Associate Professor, Civil and Environmental Engineering at Rice University
cohan@rice.edu

Daniel Cohan is an Associate Professor of Civil and Environmental Engineering at Rice University and a member of EPA’s Board of Scientific Counselors climate change subcommittee. He received a B.A. in applied mathematics from Harvard University and a Ph.D. in atmospheric science from Georgia Tech. Dr. Cohan is a recipient of a National Science Foundation CAREER award and served as a Fulbright Scholar to Australia. He is the author of more than 50 peer-reviewed publications, 75 popular media articles, and the book Confronting Climate Gridlock: How Diplomacy, Technology, and Policy Can Unlock a Clean Energy Future (Yale University Press, 2022), which was chosen by the Financial Times as one of best new books on climate and the environment.

Anshumali Shrivastava ( 11 - 11:45 a.m. )

Anshumali Shrivastava
Associate Professor of Computer Science, Electrical and Computer Engineering and Statistics at Rice University
anshumali@rice.edu

Philip Andre Witte ( 1:15 - 2 p.m. )

Philip Andre Witte
Senior Researcher at Microsoft Research
pwitte@microsoft.com

Philipp Witte is a Senior Researcher at Microsoft Research for Industry (RFI) with an interest in numerical simulations, parallel programming, and applications of machine learning to scientific computing. He received his Bachelor and Master of Science in Geophysics from the University of Hamburg and his Ph.D. in Computational Science and Engineering from the Georgia Institute of Technology. During his Ph.D., he was a member of the Seismic Laboratory for Imaging and Modeling under the supervision of Felix J. Herrmann, where he authored and contributed to open-source projects such as the Julia-Devito Inversion Framework (JUDI), InvertibleNetworks.jl and Devito. At Microsoft, Philipp is working on abstractions for running HPC workloads in the cloud, as well as surrogate modeling and optimization using large-scale deep learning and its application to carbon capture and storage.

Scientific AI for CCS: Solving PDEs for industry-scale carbon storage problems with deep learning
Solving partial differential equations with deep learning makes it possible to reduce simulation times by multiple orders of magnitude and unlock scientific methods that rely on large numbers of sequential simulations, such as optimization and uncertainty quantification. One of the big challenges of adopting scientific AI for industrial applications such as reservoir simulations is that neural networks for solving large-scale PDEs exceed the memory capabilities of current GPUs. In this talk, we discuss strategies for scaling scientific AI to commercial-scale problems and show how AI-driven PDE solvers can enable new downstream tasks in renewable energy scenarios. Specifically, we demonstrate how we can build a large-scale AI simulator for modeling subsurface CO2 flow in a real-world carbon capture & storage (CCS) scenario and use it to determine the optimal well placement for CO2 injection.

Suvinay Subramanian ( 2 - 2:45 p.m. )

Suvinay Subramanian
Senior Software Engineer at Google

Tuesday, April 18, 2023

Erik Skjetne ( 8 - 9 a.m. )

Erik Skjetne
Chief Digital Officer at Northern Lights
Erik.Skjetne@norlights.com

Karthik Kashinath ( 9 - 9:45 a.m. ) [Track 1]

Karthik Kashinath
Principal Scientist and Engineer at NVIDIA

Kevin See ( 9 - 9:45 a.m. ) [Track 2]

Kevin See
VP Product Strategy at Geminus.AI
kevin@geminus.ai

Kevin is a recognized leader in emerging technology strategy and has advised the world’s leading industrial companies on their innovation and digitalization plans. As the VP of Research and Digital Products at Lux Research, he worked across energy, chemicals & materials, and electronics industries. After completing his Ph.D. in materials science and engineering at Johns Hopkins University, Kevin spent time at the Lawrence Berkeley National Lab.

Multimodal AI for the Energy Transition
The industrial world has spent considerable resources applying data-driven AI to managing complex assets. While there have been some isolated successes in areas like predictive maintenance, there remains a large gap to reaching the ability to truly optimize processes for both productivity and carbon footprint. We will present our approach to leveraging existing modes of information like physics-based simulation, along with sparse measurement data to build models that truly reflect the behavior of these systems, while giving insights in real-time. We’ll show examples of how this enables new levels of optimization and the next generation of digital twins.

Istvan Szunyogh ( 10:15 - 11 a.m. ) [Track 1]

Istvan Szunyogh
Professor of Atmospheric Sciences at Texas A&M University
szunyogh@geos.tamu.edu

Istvan Szunyogh is a Professor of Atmospheric Sciences at Texas A&M University. He is an expert in atmospheric dynamics, numerical weather prediction, data assimilation, and the application of machine learning to atmospheric modeling. He was a co-author of the Local Ensemble Transform Kalman Filter (LETKF) data assimilation algorithm, which has become one of the most widely used data assimilation techniques for both operational and research applications around the world. His current research focuses on the development of modeling approaches that combine physics-based numerical techniques, machine learning, and data assimilation. He received his Diploma in Meteorology from the Eotvos Lorand University, Budapest, Hungary in 1991 and his Ph.D. in Geosciences from the Hungarian Academy of Sciences in 1994. He moved to the United States in 1996, where he held different research positions at the Massachusetts Institute of Technology, National Center for Atmospheric Research, National Centers for Environmental Prediction, and the University of Maryland, before taking his current position at Texas A&M University in 2009.

A Hybrid Approach to Atmospheric Modeling
This talk describes a hybrid modeling approach that combines machine learning (ML) with a physics-based numerical model of the atmosphere. The potential of the hybrid approach is demonstrated with an implementation on a global circulation model. It is shown that the ML component leads to the largest forecast improvement for the atmospheric state variables (e.g., humidity, temperature, wind) near the earth’s surface. It is also demonstrated that the hybrid approach provides a flexible framework to introduce (ML-based) prognostic model variables that are not included in the physics-based numerical component, for instance, to better represent air-sea-land interactions. Finally, it is argued that these demonstrated flexibilities of the hybrid modeling approach could be highly beneficial for the high-resolution prediction of the wind near the earth’s surface.

Umut Ozdogan ( 10:15 - 11 a.m. ) [Track 2]

Umut Ozdogan
Managing Director, Sustainability, Corva.ai

Umut Ozdogan is the Group Director for Commercial and Sustainability at Corva. At Corva, Umut leads the firm's sustainability and lower carbon technology initiatives. Umut has 20+ years of energy industry experience from 150+ energy assets around the world. Previously, Umut served as the Vice President for QRI Group, where he led the company’s growth initiatives in 9 countries and execution of $250 Million asset transactions. Before QRI, Umut worked for Chevron for 14 years in various leadership, engineering, and operations roles. Umut is a member of the Sustainability Accounting Standards Board Alliance. Umut serves as the Managing Director for Tiburon Resources, a boutique energy technology advisory firm that specializes in lower carbon & technology initiatives. Umut holds a B.Sc. in PE from METU, Masters in PE from Stanford University, and an MBA from the University of Southern California.

Haibin Di ( 1:30 - 2:15 p.m. ) [Track 1]

Haibin Di
Senior Data Scientist at Schlumberger
HDi@slb.com

Haibin Di is a Senior Data Scientist in Digital Subsurface Solutions at SLB based in Houston, working on deep learning solutions for subsurface interpretation with applications to oil & gas field exploration and new energy site characterization. He has more than 30 scientific articles in journals, 50 conference proceedings and abstracts, and published two book/book chapters. Prior joining SLB in 2018, Haibin was a postdoctoral fellow at Georgia Institute of Technology from June 2016, right after completing his Ph.D. degree in geology from West Virginia University. Haibin is a member of the SEG Research Committee.

Deep learning applications for offshore windfarm site characterization
Developing an offshore wind farm requires robust characterization of the near subsurface soils for turbine foundation design, construction, and monitoring, which faces with many challenges related to seafloor topography mapping, shallow geohazard detection, structure interpretation and modeling, soil type analysis, geotechnical parameter estimation and so on. This talk revisits these existing challenges from the perspective of pattern recognition, investigates the feasibility of deep learning in resolving them, and proposes an integrated workflow of great potential in accelerating the process of windfarm site characterization. Its values are demonstrated through applications to the Dutch wind farm zone for accelerating three major tasks, including picking multiple major horizons, mapping the seafloor topography, and estimating the essential geotechnical parameters.

David Thul ( 1:30 - 2:15 p.m. ) [Track 2]

David Thul
Founder at Geolumina (www.geolumina.ai)
dthul@geolumina.ai

David is a petroleum geologist that builds AI to augment subsurface workers. He started using data science to support exploration and production work 15 years ago. In his career he has run a multi-million dollar upstream research group, raised and deployed PE capital for oil and gas exploration, marketed and sold private drilling deals, and most recently bootstrapped Geolumina from initial concept to implementation at super major oil companies. He holds a BA and MS in Geology and is ABD on his PhD from Colorado School of Mines.

AI for the Digital Subsurface: Extract, Structure, Search, Interpret:
No matter the energy mix, the subsurface and its resources are a critical building block of modern society. Despite there being more aspects to the subsurface value chain than ever before, there are 30% fewer subsurface workers today than just 5 years ago. Scarcity of the subsurface workforce means that worker augmentation is required to achieve ambitious decarbonization goals while delivering necessary resources. Geolumina has spent the last 3 years using computer vision and natural language processing to create artificial intelligence that sees like a geoscientist. In that time, we have developed a core principle:

Automate anything that can be automated; Don’t automate anything that can’t be automated.

In this talk I will cover strategies and approaches that have led to surprising successes and failures in creating augmentative artificial intelligence for exploration and utilization of subsurface resources in oil, gas, CCUS, and geothermal.

Mohamad Nasr-Azadani ( 2:15 - 3 p.m. ) [Track 1]

Mohamad Nasr-Azadani
Principal Data Scientist at Accenture

Niven Shumaker ( 2:15 - 3 p.m. ) [Track 2]

Niven Shumaker
Digital Innovation Advisor and Data Scientist at Schlumberger
AShumaker@slb.com

Niven Shumaker is currently Data Scientist and Innovation Advisor on the Global Innovation Factori Team at SLB. He has previously consulted for an AI-based exploration start-up and Sandia National Labs. Niven spent 15 years at Noble Energy in various positions including Tamar Asset Manager, Sr. Finance Advisor, Geophysical Advisor and was a key contributor to high impact deep water discoveries in the US Gulf of Mexico, West Africa, and the Eastern Mediterranean. Niven has published numerous papers and teaches short courses on geophysics, unconventionals, pore pressure and innovation. Niven holds geophysics degrees from UC Santa Barbara and Virginia Tech, has studied geotechnical engineering at the University of Iceland, and holds an MBA with energy emphasis from Rice University, where he was awarded the Jones Scholar Award for finishing in the top of his class.

Seismicity induced by fluid injection and extraction is a widely observed phenomenon. These earthquakes can exceed magnitudes of M 6 and have the potential to impact on the containment, infrastructure, and public perceptions of safety at CO2 storage sites. This talk focuses on AI-enabled components of integrated “seismic to simulation” storage site characterization that cut cycle time, lower evaluation costs, and improve site characterization accuracy. Specifically, AI and ML methods are used to extract key geologic structures from reflection seismic images, assist in subsurface property modelling and dynamic history matching, and in developing seismic hazard maps.