Simulation modeling of distribution networks – network adaptability – part 1
The concept of a simulation model – the essence of management system dynamics models
In logistics and production networks oriented towards comprehensive customer order fulfillment within an acceptable timeframe, subcontracting tasks is an opportunity to meet these requirements. Research on subcontracting in networks of cooperating enterprises is very broad and covers many research trends. Chen and Li (2008) proposed a decision-making model involving subcontracting and job scheduling. In their proposed model, production operations, methods of order fulfillment, and order acceptance from a group of customers are determined at the beginning of the scheduling period. The authors propose the introduction of rankings to determine which orders must be prioritized and which orders should be outsourced to subcontractors. In their research, they focus on homogeneous resources of the base organization and its partners. The authors analyzed the results of subcontracting options in the context of improving the comprehensiveness of order fulfillment. The research concept presented in this article uses some of the determinants proposed by Chen and Li (2008) as well as the concept of setting subcontracting priorities. The subject of the research is those events where the integrator receives an order exceeding the capacity of its own resources within a set period of time. In this sense, it is a reference to the scheduling model with subcontracting. If companies outsource entire operations that cannot be performed within their own resources to subcontractors, uncertain supply chain needs are reduced by the risk diversification effect, which relates to the need to build network relationships within a given point in the supply chain. The authors also identified significant factors that are taken into account when making subcontracting decisions: production capacity, production costs on time, customer requests, availability of subcontractor resources, their production costs, and delivery lead time between the base enterprise and the subcontractor. The subcontracting model is therefore important if a company needs to optimize trade-offs for these factors.
Due to treating distribution networks as complex adaptive systems, one of the methods emphasized in the CAS methodology – Management System Dynamics – was used in the simulation modeling. The research concept is focused on significant aspects related to the problems of adapting to environmental changes from the perspective of a supply chain link responsible for synchronizing the demand and supply side at the distribution level. The degree of adaptability of enterprises in the distribution sector is influenced by endogenous factors related, among other things, to possessed resources and established relationships, as well as exogenous factors related to the micro- and macro-environment. The concept considered the possibility of simulation modeling using the System Dynamics method, taking into account endogenous and exogenous factors.
System dynamics in management sciences was initiated by Forrester (1958) and subsequently popularized by Senge (1990). In their works, they noted that the structure of feedback loops generates patterns of behavior. This causes difficulties in finding cause-and-effect relationships. The effects of actions often occur after a lapse of time, which also makes it difficult to identify the causes. System dynamics consists of its change over time. Therefore, when studying the dynamics of a distribution network as a complex adaptive system, one must answer questions similar to those posed for a traditional management system:
-
does the system change and are these changes captureable (e.g., two states of the system in a specific time interval),
-
is the shift between these states characterized by constant, uniform movement or not,
-
is it predictable or not.
Management system dynamics uses a specific method of describing a system based on two categories: flow, level. Levels represent state variables that are observable quantities of the system, and flows represent the system’s actions that cause changes in level values. The mentioned description categories are essential for creating the microstructure of feedback loops and thus the entire system. In practice, it is useful to introduce a certain additional category of description called an auxiliary variable. These variables represent intermediate stages of the process of determining flow intensity values according to adopted rules (decision policies) that transform information about the state of the system and its environment. Decision rules describe the ways of making decisions that control flows in the system, that is, they show what information about the state of the system and its environment generates decisions that cause actions in the system. In classical MSD, it is assumed that decisions in the system are made periodically, that is, each time after a specific time interval has elapsed. Decision-making processes, and thus their effects (changes in flow intensity values), are therefore continuous in nature. The form of decision rules is conditioned by, among other things, socio-economic, technical, and organizational factors.
The article comes from the journal “Logistyka” 3/2011.