The DECISION project aims to develop:

• An optimization 'platform' which integrates new and existing optimization technology in a problem solving environment suitable for tackling complex design optimization problems. Emphasis is placed upon the exploitation of HPCN environments. NAG is responsible for developing a commercial version of the platform, based on its IRIS Explorer visualisation and application development system.
• A 'database' incorporating details of test problems from the end-user partners, specialized visualization software, and guidelines for optimization procedures.

Three major stages may be identified in the DECISION project:

1. Specification and design.
2. Development of optimizers, including: traditional deterministic, gradient-based; non-smooth; stochastic (genetic algorithms and simulated annealing); multicriteria (including Pareto and Nash equilibria concepts).
3. Development of the optimization platform. In addition to the actual platform development, this stage also involves its application to a number of end-user problems, the development of commercialised products and the formulation of strategies for product marketing.

The expected benefits of products deriving from the project include the potential for a significant reduction of R&D costs in complex engineering projects. The ability offered by the optimization platform to hybridise a range of deterministic and stochastic optimization algorithms aims to ensure that problems can be solved with a high level of accuracy and robustness.

The end-user problems are an important aspect of the project as they will not only inform the design of the optimization platform, providing a test bed during the development phase, but will also demonstrate the effectiveness of the platform in tackling 'real world' problems. The results from, and demonstrations of, these test problems will be invaluable in marketing the commercial products derived from DECISION.

The project offers many possible exploitation paths. The main commercial product will be a general purpose design optimization environment, but the technology developed, combined with the existing application-building features of IRIS Explorer, has the potential to to provide products specialized towards particular markets. In future, markets could extend beyond the engineering design optimization field targeted by the project, to encompass areas such as finance and operations research. Additionally, NAG's long-established status as a supplier of numerical library software provides a separate exploitation route for the optimization software developed during the project.