Jon has over 15 years of experience in the oil and gas industry working with NOV. He has held senior roles in global manufacturing and supply chain, Corporate Development (M&A), and now leads the Advanced Analytics organization for the Intervention and Stimulation Equipment business unit of NOV. Jon holds a bachelor’s degree from the University of Texas at Austin, an MBA from Rice University, and a Master’s of Data Science from Northwestern University.
Connectivity limitations, latency requirements and the cost of transmitting data from the field has motivated increased adoption of accelerated edge computing devices capable of processing large volumes of data at the source. These low-cost, low-power GPU, TPU and FPGA-accelerated devices allow compute-intensive tasks to be performed at the edge, without relying on internet connectivity, to provide insights and even automate tasks. These devices are being deployed for ML-based equipment optimization and predictive failure analytics algorithms to run in real-time at the edge. Additionally, real-time inferencing using convolutional neural networks on video data captured from cameras deployed across the energy’s industry’s edge – including wellsites, service trucks and manufacturing facilities – has tremendous implications for field safety, operations intelligence gathering, and improved product and service quality. This presentation will share technical challenges, key learnings and how Baker Hughes is leveraging accelerated edge computing to enable the digital transformation of the oilfield.
Jeff is responsible creating and implementing strategic direction in Cyber-Physical Systems across Baker Hughes – systems which integrate emerging technologies such as computer vision, accelerated edge computing, advanced sensors and augmented reality. Based out of the Baker Hughes Energy Innovation Center in Oklahoma City, Jeff leads an interdisciplinary team of senior technologists and coordinates with technology partners to deliver new cyber-physical solutions for the energy industry.
Jeff holds a PhD in Materials Engineering from the University of Texas at Austin and a bachelor’s degree in Mechanical Engineering from Oklahoma State University. He has published over twenty technical publications and filed over ten patent applications, with technical and leadership experience in a variety of areas including artificial intelligence, energy storage, manufacturing and materials development.
Optimization of the end-to-end oil and gas value chain, where multiple plants, processes and assets are interdependent, has been a complex challenge for the upstream oil and gas industry. Canada’s largest integrated oil and gas producer set out to apply the latest artificial intelligence (AI) and data science methodologies to improve production optimization within their upstream operations facilities. The goal of the project was to optimize across siloes, flag upsets early for timely response and identify and action opportunities in real-time. The scale of the project included a 130 Km2 area with 35 individual plants which included over 134,000 individual tags operated by thousands of front-line operators, engineers and supervisors who spanned 12 individual business units.
Over one hundred machine learning models in a multi-layered approach where applied to bring value to the complex problem in both Normal and Upset conditions. This ‘systems of systems’ approach included advanced AI models for a predictive systems layer - including predictive and adaptive mass balance models - an optimization layer with end-to-end optimization models and an opportunities and process layer for opportunity identification including 58 variations of process upset flagging models.
In order to truly scale production optimization, a dynamic optimization engine connected to business unit specific optimizers was implemented to rapidly generate production schedules and create a new plan in less than 10 minutes to maximize overall plant throughput and production performance. By continuously monitoring production every two to three minutes, the solution identifies gaps between current and most effective and achievable – not theoretical - variables, providing recommendations to maximize business objectives such as production volume, quality, inventory levels, profitability etc.
Plant operators are placed in full control of process upset management and opportunity awareness enabling them to predict and minimize plant upsets and take action to optimize processes and quality where required. The solution currently predicts upsets 20-25 minutes before with more than 80% accuracy. Opportunity awareness further helps operational decision makers action opportunities to further improve product quality and minimize energy use.
Dariusz Piotrowski leads the strategy and development of global AI (Artificial Intelligence) solutions in natural resources (oil, gas and mining). Dariusz specializes in large transformational projects focussed on business and operational process optimization, machine learning and advanced AI models. Dariusz has over 27 years of technology and consulting experience working with senior leaders within some of the world's largest natural resources companies to help them realize business value through fusing together advanced technologies, data science, applied R&D and agile methodologies and practices to transform business processes and performance. Dariusz holds an architecture and civil engineering degree from Warsaw University of Technology in Poland and an MBA from the Richard Ivey School of Business in Western Ontario.
Jim Claunch is a digital visionary with over 35 years of experience in the energy sector. He is an expert and proven thought-leader in digital transformations of the upstream sector. Jim has extensive experience engaging industry leaders to develop a coordinated, industry-wide approach. His experience includes organizational change and development, leadership, people development and workforce planning.
He has held various senior executive positions including CFO, CIO, SVP, and VP across multiple companies and industries. Included in Jim’s 35 years of experience are 14 years of living and working outside of the US.
A grand challenge in machine learning is the development of computational algorithms that match or outperform humans in perceptual inference tasks that are complicated by nuisance variation. For instance, visual object recognition involves the unknown object position, orientation, and scale, while speech recognition involves the unknown voice pronunciation, pitch, and speed. Recently, a new breed of deep learning algorithms have emerged for high-nuisance inference tasks that routinely yield pattern recognition systems with super-human capabilities. Similar results in language translation, robotics, and games like Chess and Go plus billions of dollars in venture capital have fueled a deep learning bubble and public perception that actual progress is being made towards general artificial intelligence. But fundamental questions remain, such as: Why do deep learning methods work? When do they work? And how can they be fixed when they don’t work? Intuitions abound, but a coherent framework for understanding, analyzing, and synthesizing deep learning architectures remains elusive. This talk will discuss the important implications of this lack of understanding for consumers, practitioners, and researchers of machine learning. We will also briefly overview recent progress on answering the above questions based on splines functions and computational geometry.
Richard G. Baraniuk is the Victor E. Cameron Professor of Electrical and Computer Engineering at Rice University and the Founding Director of OpenStax. Dr. Baraniuk’s research interests lie in the areas of signal, image, and information processing and include machine learning and compressive sensing. He is a Fellow of the American Academy of Arts and Sciences, National Academy of Inventors, American Association for the Advancement of Science, and IEEE. He has received the DOD Vannevar Bush Faculty Fellow Award (National Security Science and Engineering Faculty Fellow), the IEEE Signal Processing Society Technical Achievement Award, and the IEEE James H. Mulligan, Jr. Education Medal, among others.
The big bang of Artificial Intelligence was in 2012, when a student in Canada combined the three necessary ingredients: big data, deep neural networks, and GPUs. Since then, we've seen a Cambrian explosion of deep learning algorithms capable of tackling a wide range of problems from cancer detection to autonomous drilling. However, the history of GPUs and O&G started several years earlier with the parallelization of seismic processing codes.
This talk discusses the past, present and future of GPU computing. We will dive deep into the current hardware and software toolkits from NVIDIA that enable the training and deployment of ML/DL models in the cloud, at the edge, and in the datacenter.
As a Global Industry Director, Marc has responsibility for NVIDIA’s energy strategy and business development. Before joining NVIDIA, Marc spent 13 years with Halliburton, where he held positions in Commercial and Strategic Alliances, Technology Operations, Customer Financial Services, and Corporate Development. Prior to Halliburton, Marc worked for Silicon Graphics, Inc. (SGI) where he held a variety of sales and business development roles in the energy vertical. Marc has an MBA from Rice University in Texas and a Master of Science in Professional Development and Leadership as well as Bachelor of Science in Marketing from Winona State University in Minnesota.
Mark Anderson advises companies and boards on project specific matters. He provides consulting services with emphasis on drilling technology. Previously Mark worked for Shell for 37 years with Shell in a variety of Well Engineering and Well Delivery Operations positions, with both technical and management responsibilities. His position before leaving Shell was as General Manager Drilling Mechanics Technologies. Mark is also Chair of SPE’s Drilling System Automation Technical Section and sits on the Center for Houston’s Future – Future of Energy O&G Working Group.
Mohamed Sidahmed, Ph.D., is subject matter expert in machine learning & AI, data-driven modeling and optimization with both theoretical and applied skills. He is the Machine Learning and AI R&D Manager at Shell’s Data Science CoE, where he is leading a multidisciplinary research group with passion for delivering innovation and excellenceHe has numerous publications and book chapters in the areas of pattern discovery, deep learning representation, and modeling & reasoning across multiple domains.
Philip graduated with a B.Sc. in Earth Sciences and an M.Sc. in geophysics and computer science in Geneva, Switzerland in 1979. After working with one of the top multinationals on a global technology transition in 2008-2009, he has focused on product strategy and the alignment of data, science, AI and ML technologies and visualization to the long-term objectives of oil and gas operators. He now works for the industry consortium Energistics in Houston, where he is responsible for the valuation strategy to drive adoption of digital data exchange standards. He is a member of SEG, SPE and EAGE.
Liz Percak Dennett is a passionate technologist 10+ years of experience and deep domain expertise in oil and gas geology. Through this interdisciplinary skill set, she has demonstrated success translating academic advances to scalable industry solutions in a wide range of use cases. Liz is currently a Solutions Architect at AWS where she is focused on bringing cloud-computing based workflows to the Energy sector. Before that, Liz worked as a geologist for Hess Corporation across a wide range of functional groups and teams. Liz holds a BS in geology from the University of Alaska Anchorage and an MS and PhD in geoscience from the University of Wisconsin-Madison.
Dr. Milos Milosevic is Senior Director Technology at Landmark. He is responsible for the technology development and product management supporting Landmark’s Exploration and Reservoir Characterization, Information Management and Platform Technology, Drilling and Completions, Production and Reservoir Engineering product lines. He was most recently the Director of Technology for the Production Group at Halliburton. Milos holds B.S. and M. S. degrees in electrical engineering from Illinois Institute of Technology; MBA from Rice University in Houston; and doctorate in electrical engineering from The University of Texas at Austin. Dr. Milosevic is the author of several patents and publications in peer-reviewed technical journals and conferences.
John Thurmond is a leading research geologist for Equinor, and has spent his career working at the boundary of computing and geology. He received Ph.D. in Geosciences from the University of Texas at Dallas. He joined Norsk Hydro in 2005, and has held a variety of positions within what is now Equinor including geology discipline lead for one of Equinor’s largest production assets and as a exploration manager for Mexico during their first competitive licensing rounds. He has spent the last few years in research working on finding new technologies that will allow radical change within the oil and gas industry, including the application of machine and deep learning to the interpretation of seismic data.
Jaidev Amrite leads product definition for SparkCognition’s Natural Language AI, DeepNLP.
DeepNLP enables Subject Matter Experts in O&G, Aerospace, Finance and Telco, to automate analysis of natural language content without requiring expertise in Data Science or Programming. Before SparkCognition, Jaidev led product strategy for Web, Cloud and Test Data Analytics at National Instruments. Jaidev earned his master’s degree in Electrical and Computer Engineering from Georgia Tech.
Dr. Tailai Wen is a Lead Data Scientist at Arundo Analytics, where he is overseeing the data science projects in one of Arundo’s largest global accounts and leading two open-source projects. He has 10+ years of experience connecting data-driven technologies with industrial sectors, including oil and gas, transportation, chemicals, and renewable energy. Dr. Wen has a Ph.D. in computational mathematics and an M.S. in statistics from Stanford University, and a B.S. in mathematics from Tsinghua University.
Dr. Hector Klie is an experienced computational and data scientist focused on developing physics-informed AI solutions for multiple engineering and geoscientific applications in Oil & Gas. Dr. Klie is co-founder and Chief Executive Officer of DeepCast.ai He has published over 80 papers in the areas of sparse linear solvers, production forecasting, field optimization, uncertainty quantification, high performance computing, reduced order modeling and machine learning. Dr. Klie has patented 5 inventions in the areas of data analytics for automated drilling and parallel physics-based solvers. Dr. Klie completed his Ph.D. in Computational Science and Engineering at the Dept. of Computational and Applied Mathematics at Rice University, 1997, and a Master Degree in Computer Science at the Simon Bolivar University, Venezuela, 1991.
Joshua Adler is the Founding Chief Executive of Sourcewater, Inc., the water intelligence platform for the energy ecosystem, which he conceived in MIT’s Energy Ventures program. Sourcewater’s digital platform for water intelligence shows where oilfield water comes from and goes to on the surface and in the subsurface utilizing advanced data gathering and integrity methods, satellite imagery analytics, tens of thousands of mobile location sensors and geoscience research.
Steve is a Digital Business Leader who solves business problems through the effective use of data and analytics. Over the course of his career in oilfield services, consulting, and technology he has developed and implemented strategies to take on the biggest issues that companies must overcome in order to succeed in digital transformation of their businesses or industries. Now at C3.ai, Steve works with clients across several domains to design and deploy machine learning applications at enterprise scale to solve previously unsolvable problems.
Dr. Ligang Lu is a Principal Science Expert and Principal Machine Learning Scientist in Machine Learning and AI R&D at Shell Technology Center, Houston, TX. Ligang has authored or coauthored over 80 peer-reviewed publications and holds 41 issued patents, some of them have generated high business value and/or have been widely by researchers and practitioners. He was the chair of IBM Research Signal Processing Professional Interest Community and served as the general co-chair, editor, organization committee member, (special) session chairs of many international conferences and workshops in his research areas. Ligang is an Adjunct Professor at Texas A&M University.
Mehrdad Gharib Shirangi is a Staff Data Scientist at Baker Hughes Oil Field Services where he is leading various projects on prescriptive data analytics, automated machine learning, and digital transformation. Shirangi holds a PhD degree from Stanford University, an MS degree in Petroleum Engineering from University of Tulsa, and BS degrees in Mechanical Engineering and Petroleum Engineering from Sharif University. He has published over twenty technical papers in various subjects including digital twins and digital transformation for drilling and completion, closed-loop reservoir optimization, and reservoir forecasting.