Statoil uses intelligent agents for responsiveness

Oil trading and operations management support software using intelligent agents is being developed for Norwegian Statoil by Cambridge Consultants. Intelligent agents – autonomous software applets that encapsulate knowledge – will be applied to help solve optimisation, planning and process control issues in Statoil’s trading and operations areas. Among expected benefits in the trading area are up-to-date information on-tap, enabling the company to respond faster to market changes. Statoil says that the increased speed of the operation will allow it to improve effectiveness and continually modify business tactics, enabling it to maximise revenues, while reducing decisions made on out-dated information. First applications are scheduled to start going live in April. Statoil also plans to explore the use of intelligent agents in other areas – including gas operation and remote monitoring of production facilities. Statoil says it chose the JACK intelligent agent development toolkit from Agent Oriented Software (AOS) because of AOS’s track record in knowledge engineering and support availability from Cambridge Consultants. Each agent has an internal structure based on Beliefs, Desires and Intentions (BDI), which can be used to make systems behaviour-driven or provide attributes such as decision support. AOS’ JACK language, based on Java, is a framework that allows developers to build, debug and run intelligent agents.