An architecture for multi-dimensional temporal abstraction supporting decision making in oil-well drilling

Abstract

We have developed an online decision support system that advice drilling engineers online based on data streams of real-time rig site measurements. This is achieved by combining multi-dimensional abstraction for recognizing symptoms and case based reasoning. Case-based reasoning compares the current situation in the well with past situations stored in the case base that contains advices for how to solve similar problems. The architecture is described in detail, and an example is presented in depth as well as results of commercial deployment. © Springer-Verlag Berlin Heidelberg 2013.