A Framework for Introducing Data Analytics and Machine Learning to Petroleum Engineering Undergraduates

Abstract It is the age of data; data is everywhere! Digital transformation of industries is fueling an exponential growth in global datasphere, enabled by advances in data-measuring smart devices, Internet of Things and cloud computing. The oil/gas sector is not left out of the data revolution. A recent report indicated that about 80,000 sensors generate 2 terabytes/day data on an average offshore platform. The need to transform oilfield data into insights for optimal decision-making is opening up career frontiers in petroleum data science. An increase in demand for talents in petroleum data science is in view. In anticipation of this, the Data Science and Engineering Analytics Technical Section of the Society of Petroleum Engineers, African Region is collaborating with the academia, towards producing graduate engineers possessing domain knowledge and data mining skills. As part of the collaboration, we now present a framework for the incorporation of data science education into petroleum engineering undergraduate programs. In line with Outcome-based Education approach, a set of learning outcomes for a proposed petroleum data analytics and machine learning course is presented, first. These learning outcomes are framed as skills and competencies expected of students at course completion. Thereafter, assessment rubrics and use case examples are mapped to each learning outcome. These rubrics serves as evidential basis for ascertaining that the skills have been acquired. In completing the curriculum design, a sequence of lesson and lab modules are enlisted as contents over which learners must gain mastery in order to meet the criteria of the assessment rubrics. Furthermore, various elements that make up an integrated learning experience are presented. These include open-access learning resources and toolbox, open-access data sources, instructional strategies, capstone project ideas, research prospects, and industry support (mentorship and advisory). The future prospects of this initiative are also discussed. Some elements of this framework, as presented in this paper, are currently being implemented in a quasi-classroom setting; preliminary performance reports are included in this paper.