Improvement of the Decision Making Process Based on Queueing Theory

Abstract

In this paper, some organization that is considered as a queueing network consisting of n number of queueing systems (hereinafter referred to as “QS”), which are units of the organization, is to be investigated.

At the first stage, it is proved that the flow of incoming questions (so-called problems that need to be solved) is a random variable having the Poisson distribution. This allows to determine the intensity of incoming questions to the employee.

The second step of the study is to prove that the decision-making time (processing of incoming questions) is a random variable and has an exponential distribution. From this proof follows the intensity of decision-making.

At the third stage, a network of queueing systems, which consists of n QS and has some source of questions, is considered. Further, a matrix of probabilities of the arrival of questions from one QS to another is constructed, i.e. transfer matrix. The obtained data are subject to the analysis, the conclusion about each QS and its efficiency is drawn.

Step four is the definition of the characteristics of the QS network. At this stage, an intensity of the questions flow to each system is determined, as well as the intensity of decision-making on each issue.

Knowing the intensity of incoming questions, as well as the intensity of decision making, the optimal number of employees necessary for the effective operation of the organization as a whole is defined.

In completion of work the conclusion about the application of queueing theory in the process of making managerial decisions, which will allow to determine in which link the problem arises, who exactly slows down the process of solution of questions is made. The approach considered in this work will allow to avoid negative results of the organization's work by identifying problem areas at the decision-making stage, and not at the final stage – summing up.