Method for optimal capability-based R&D portfolio selection

This paper presents a method to select a set of risky R&D projects toward a long-term capability goal. Achieving a long-term capability enhancement in a modern complex system often requires simultaneous improvements in multiple subsystems. Choosing an optimal investment portfolio from a set of risky R&D investments becomes challenging when different combinations of improvements may result in the same capability enhancement. In the presented method, a collection of root portfolios is identified, each of which achieves the target capability. The likelihood of success and cost (development and production) for each project within in the portfolio is then determined. Generally, prospective projects will be included in multiple root portfolios. An optimal R&D portfolio, composed of one or more root portfolios, can then be calculated to minimize portfolio cost (development, production, or both) for a given likelihood of success and a small set of dominant portfolios may occur. The method inherently captures the interdependence and trades that occur during portfolio selection. The method is demonstrated on a notional R&D portfolio.