An industrial dynamics model for start-up management

Abstract

Efficient start-up management can result in considerable productivity increases compared to improvements possible in `steady-state' activities. Provided adequate targets are set and updated in accordance with realistic productivity expectation, management can interact rapidly in order to achieve optimum performance. Such a management process is dependent on the availability of suitable methods of predicting future productivity and estimating present productivity in the presence of random fluctuation due to many causes. This paper develops a predictive start-up model which performs both these functions. At each observation time, observed productivity is compared with predicted productivity as estimated from previous data, the difference being used to update the model parameters in a manner analogous to exponential smoothing. The model is applied to a number of start-up problems and is shown to track the parameters in an acceptable fashion. Long- and short-term productivity predictions resulting from the model are shown to be a useful management aid.