Natural language generation specialist Arria (OITJ June 2014) has announced Polus, a joint venture with IBM that applies the latter’s Watson artificial intelligence engine to leak detection and repair in chemical plants and oil refineries. Polus uses NLG to provide operators with a textual summary of a plant’s current fugitive emissions.
Watson’s Jeopardy-winning question-answering faculty has been interfaced to the large amount of legislative documentation that governs fugitive emissions (an area where non-compliance can mean fines). Watson then generates plain English reports that answer the operator’s query, possibly asking for more input from the operator. The system also generates reports automatically from monitoring data.
Arria showed off Polus at the recent IBM ‘World of Watson’ event in New York. Other uses of Watson include Perficient’s Incident-Insight and EnGeo’s GoFetchCodes. IBM is reported as planning a $1 billion investment in its Watson line of business.
© Oil IT Journal - all rights reserved.