یک روش دقیق برای برنامه ریزی فن آوری های جایگزین در پروژه های "تحقیق و توسعه"

A fundamental challenge associated with research or new product development projects is identifying that innovative activity that will deliver success. In such projects, it is typically the case that innovative breakthroughs can be achieved by any of several possible alternative technologies, some of which may fail due to the technological risks involved. In some cases, the project payoff is obtained as soon as any single technology is completed successfully. We refer to such a project as alternative-technologies project and in this paper we consider the alternative-technologies project scheduling problem. We examine how to schedule alternative R&D activities in order to maximize the expected net present value, when each technology has a cost and a probability of failure. Although a branch-and-bound algorithm has been presented for this problem in the literature, we reformulate the problem and develop a new and improved branch-and-bound algorithm. We show using computational results that the new algorithm is much more efficient and outperforms the previous one.

The development of complex and innovative products is characterized by much uncertainty. In order to deal with this uncertainty, it has been suggested that research and development (R&D) projects should pursue multiple alternative solutions for developing the new products (see, for instance, [1] and [2]). The scheduling of these attempts, hereafter referred to as alternatives, is crucial for increasing the likelihood of successfully developing a product, minimizing development time and obtaining revenues as early as possible. Consider, for instance, a software development firm that has the option to develop their web services using either a traditional Java SPRING framework or the pioneering Ruby-on-Rails framework. While both might achieve a similar functionality, the traditional Java SPRING framework will take longer to develop, but is more likely to handle the expected volume of users. A similar situation happens in the formulation, delivery and packaging development phase of the pharmaceutical drug-development process in which drug developers must devise a formulation that ensures the proper drug delivery parameters. It is critical to begin looking ahead to clinical trials at this phase of the drug development process. Drug formulation and delivery may be refined continuously until, and even after, the drug’s final approval. Trials have different costs, durations and probability of success, and optimal scheduling of these trials saves a noticeable amount of money for the drug developer firm (see [3]).

In this paper, we reformulated the ATPSP as a non-linear integer programming model and proved that there is an optimal concurrent schedule for each instance of the ATPSP. Also, we developed a new branch-and-bound algorithm for the problem which relies on the concurrency property. The computational results indicate that the F-B&B algorithm is faster than FB-B&B and O-B&B algorithms but is not able to find optimal solutions in the case there are unfavorable alternatives. Also, for PTC=5, the preprocessing procedure was not efficient but becomes more effective as PTC increases. In addition, the developed upper bound has positive impact on the performance of the B&B algorithms especially for larger test instances. Since exact algorithms are not able to solve large scale instances of the ATPSP, as a future research opportunity, we recommend to develop heuristic or metaheuristic algorithms for the problem. We also suggest to develop the setting of this problem to more practical assumptions like stochastic durations for alternatives (see Creemers et al. [6]) or resource constraints.