SCENARIO-DRIVEN INFORMATION RETRIEVAL: SUPPORTING RULE-BASED MONITORING OF SUBSEA OPERATIONS

Abstract

The production systems used by the subsea petroleum industry are knowledge and information intensive. Any problem needs to be solved quickly and efficiently avoiding decommissioning or waiting for the symptoms to be escalated. This requires precise information to be supplied on-time. For this reason we have proposed rule-based moni-toring of device performance. However, covering all possible cases by rules is a labour-intensive and not trivial task. Therefore, in this paper we propose a scenario-driven information retrieval approach to complement rule-based moni-toring. The main objective is to automatically formulate a query that is sent to a vector-space model information re-trieval engine every time incomplete inference happens, i.e. when a specific case has no rules defined.