Decision Support and Monitoring Using Autonomous Systems

Abstract

A major trend in the oil and gas industry in the recent years have gone towards moving installations from the surface to the seabed utilizing existing surface based facilities or by developing new onshore facilities. Enabling this trend is an increased instrumentation of the equipment, down-hole and on the seabed. Increased instrumentation results in more data that has to be processed by human operators. Research shows that human operator information overload leads to poor decision making. To avoid information overload, we need a new breed of system where more of the information processing and decision making can be handled by computers.Over the last couple of years we have worked extensively with autonomous systems. We see autonomy as a vital property in this new breed of systems, where we have identified three scenarios that can benefit. These are:• Production and process optimization • Environmental monitoring and control • Equipment condition monitoringThe key elements of our approach are a combination of the distributed data processing capabilities of software agents with a variable autonomy approach to the interaction with the human operator. Variable autonomy enables a human operator to defer control to a system when the system state is within a given envelope. The system will escalate decision making to the human operator when it reaches its decision making boundary.The goal of our work is to make the human operator more capable of handling critical situations and prevent unwanted incidents by transforming the stream of data into information and proposed actions. The work will be focused around the transformation of data into decisions and the interaction between the human operator and the computer system.

New risk control mechanism for innovative deepwater artificial seabed system through online risk monitoring system

“…Bayesian estimator [21], Kalman filter [22], Markov Chain Monte Carlo (MCMC) [23], etc. Normally, choosing and calibrating an estimator are not trivial, as it requires a combination of theoretical and practical understanding of the estimation models as well as the mathematics and process or equipment involved [24]. The aim of the diagnostic system is to detect and help explain all the deviations.…”