IBP - Rio Oil & Gas Expo and Conference 2020
The O&G industry is being disrupted at multiple levels by changes in the global energy landscape, business and operating models, and geopolitical order. Such a confluence of elements occurs at the moment when a technological revolution is also underway. This work focuses on discussing what Artificial Intelligence (AI) means for the subsurface data analysis and G&G disciplines. Despite all the enthusiasm around the topic and the proliferation of AI pilot projects, an effective insertion of these technologies in the mainstream of G&G workflow still poses significant challenges. Some of them are related to the unique characteristics of subsurface data, such as marked data sparsity, a high degree of uncertainty, strong spatial dependence, and complicated physics background. Other challenges are more related to the business and operational context of the O&G industry, such as domain silos in data infrastructure, difficulties to assemble teams with the required mix of skills, obstacles to deploy ML and DL solutions in a timely and reliable way and availability of domain-specific AI platforms. This work will demonstrate some examples of how a cloud-native AI platform for G&G can be employed to automate the analysis of large-volume datasets at scale and to solve existing subsurface problems such as prediction of missing well log curves, seismic inversion, reservoir properties prediction, seismic interpretation and seismic data compression.
