Simulation of disassembly and re-assembly processes with beta-distributed operation times

Abstract

In the case of designing or improving industrial disassembly and repair shops, a special characteristic to be considered is the stochastic nature of disassembly, diagnosis and re-assembly times. Because of this special characteristic, the reorganisation of such work systems shall be supported by using a simulation method which considers distributed operation times. A question to be answered is the type of distribution which should be used. The problem is that during the data collection phase of a simulation project, the parameters of the often suggested Normal- or Gamma-distribution are normally not available. In contrary the parameters of the Beta-distribution are easy to get, because it could be characterised by the optimistic, the pessimistic and the most common time value. The following investigates in which way the different distributions effect the quantitative evaluation of a modelled repair shop compared with fixed operation times. Finally disadvantages and advantages of the use of different distributions in a simulation study will be discussed. © 2005 by International Federation for Information Processing.