Since the adoption of data analytics in our industry over the past few years, there has been a growing realization of the need to not see traditional physics and the new data-driven methods as dichotomous solutions. How can we overcome the problem of “small data” in reservoir management and field development by constraining data-driven models with physics? Can we generate interpretable and physically consistent insights using these hybrid methods? What practical business problems can we solve in reservoir engineering applications, where pure physics or pure data-driven models are not adequate. In this session, we will explore various aspects on business needs and impact, promising current approaches and future needs for combining data-driven and physics-based models in reservoir engineering applications.