Interactive R&D portfolio selection considering multiple objectives, project interdependencies, and time: a three-phase approach

Summary form only given. This paper describes a stepped approach to R&D project selection taking into account interdependencies, multiple objectives, and various resources. The effects of all portfolios are traced over time: That is, the levels of resource consumption and benefit generation associated with each portfolio are available for every period t within the planning horizon T and may be used as decision aids both as aggregates (present values) or separate entities. In what follows we outline the three phases of our method. In Phase 1 a screening process identifies proposals that are worthy of further evaluation, where the number of projects entering phase 2 are kept within a manageable size. Phase 2 determines all Pareto optimal portfolios. For this purpose the portfolio selection problem at this stage is framed as a multiple objective integer linear program. Phase 3, finally, aims to select that portfolio which best fits the decision-maker´s notions