A total systems approach to R&D project evaluation, selection, and resource allocation using the analytic hierarchy process with interval judgements

Research and Development (R&D) project evaluation, selection, and resource allocation problems have been of interest to both model builders and practitioners, at least since the early 1950\u27s. The nature of the problem of R&D management (time duration and the inherent uncertainty, difficulty in measuring R&D contribution, project interactions, multiple goals and conflicting alternatives) differentiate it from other organizational activities. There are many evaluation and selection models available in the literature, but unfortunately implementation of these models in organizational settings has been limited. This dissertation presents a model that uses the knowledge and experiences of managers/researchers, project records, and input from the customers of the R&D Division (business units, corporation, other organizational units, and outside customers) in making R&D project selection decisions. Important features of the model are: (a) The model ties strategic and tactical goals of interacting units to the selection process. Long-term relative influence or dominance of the goals of the interacting units was determined using the supermatrix analysis technique and then used as an input into an overall Analytic Hierarchy Process (AHP) model. (b) The model handles multicriteria and uses actual project data in its development to provide validity to its results. (c) The model demonstrates the application of the Analytic Hierarchy Process with interval judgements (AHPIJ) to R&D project selection under risk and uncertainty. (d) The model possesses the potential to improve working relationships between researchers, business units and corporate personnel; and thereby attempt to achieve organizational commitment to an effective technology. The model was tested on thirty projects considered for implementation by a division of a central research organization and is believed to provide an acceptable, if not superior, R&D project portfolio selection. It is expected that the model will be incorporated into the project selection and budgeting process of the organization