Heavier equipment, more challenging operational conditions, increased complexity in technology and integrated operations involving multiple vessels and autonomous units are trends that point in the direction of increased complexity in marine operations. At the same time, the cost of delays combined with the focus on safety creates a need for more accurate assessment of weather limitations. A decision support may rely on real-time simulator models, state estimators and acquisition of operational data. The complexity and diversity of operations and equipment makes development and implementation of such systems challenging, and there is a need for new tools and methods for further developments of systems customized for both ships and the operations performed by those. Decision support systems will provide information accommodated to the ships captain's needs at a certain time of the operation.

Knowledge needs

There is a need for standardized methods for sharing non-mission-critical operational data on board ships and for methods for taking advantage of operational data for operations, fleet management and research. This includes methodology for storage, organisation, transfer and analysis of such data. As some information is expensive and/or difficult to measure, there is a need for state estimators and mathematical models to replace such measurements. A state estimator is integration between sensor signals and mathematical models representing e.g. the ship, estimating the conditions (states) of the system. This technique is well-known and widely used in e.g. ship dynamic positioning systems. The mathematical models used in state estimators need to be based on physical models similar to those used in engineering tools, but still optimized for computational speed as they have to run in real-time and parallel to the operation itself.

Research tasks

1. Develop methods and tools which simplify integration of e.g. decision support tools with onboard ship systems. Specifically, sharing of operational data a) From ship subsystems to non-mission-critical software tools, b) From non-mission-critical software tools to ship subsystems, c) Between ship subsystems and d) Between non-mission-critical software tools.
2. Develop ship type-agnostic methods for handling large amounts of operational data, with respect to a) Efficient ship-borne and centralized storage, b) Flexible and efficient analysis and c) Efficient and secure transfer between ship and shore.
3. Develop techniques for providing decision support based on historical operational data, for example based on statistics, artificial intelligence, case based reasoning or fuzzy logic.
4. Develop methodology for developing and implementing system state observers and decision support systems based on adequate faster than real-time mathematical models of systems such as crane operations, energy system, vessel dynamics, and propulsion units. The particular challenge is to maintain the models' properties with respect to describing physical behavior of sub-systems developed in RA 2, and to transform this into useful on-line information for ship captains during the specific stage of the operation.

Develop simplified CFD methods for implementation into real-time applications like decision support systems and training simulators as described in RA 4. The Lattice Boltzmann methods (LBM) will be the basis for this development due to their suitability for parallel execution which can be used to improve the computation efficiency. High performance computing at low cost can be achieved by use of graphics processing units (GPU's) as a resource for parallel execution, and this approach will be evaluated.