Investing in Ideas in an R & D Department

Abstract

This paper reviews some literature on project appraisal and concludes that most of the methods suggested have not been used in industry and that there are good reasons why they should not be used. One of the major criticisms of project selection methods is that most ignore the constraints imposed by limited resources. Many ignore the probabilistic nature of the problems. A simulation model is developed in order that methods of project selection may be tested. The methods tested are a simple economic ranking procedure and a stochastic LP procedure which takes account of resource limits. A random selection is also tested in order that the benefits of the formal selection techniques may be judged. Where the projects available are relatively small and numerous and are unlikely to cause scheduling problems in whatever order they are undertaken, the stochastic LP model does almost as well as an economic selection rule and significantly better than a random selection. Where scheduling problems may be encountered, the stochastic LP model can yield significantly better results than an economic model which ignores resource limitations.