Academic articles

The aero-engine value chain under future business environments

Agent-based modelling is gaining popularity for investigating the behaviour of complex systems involving interactions of many players or agents. In this paper an agent-based simulation modelling technique is applied to understand the long term implications of strategy decisions for an aerospace value chain. The industry has unique elements including new business models, high levels of collaboration, long product lifecycles and long periods before positive paybacks are realised. Forecasting market conditions over this long term lifespan is inherently problematic and adds further complexity when devising a strategy. The model described includes all the major players and entities in the value chain and their interactions. Illustrative results are presented to demonstrate how the simulation approach can be used to evaluate strategy and policy decisions and their operational implications over the long term

Using AnyLogic and agent-based approach to model consumer market

In the highly dynamic, competitive and complex market environments (such as telecom, insurance, leasing, health, etc) the consumer’s choice essentially depends on a number of individual characteristics, inherent dynamics of the consumer, network of contacts and interactions, and external influences that may be best captured within the Agent Based modeling paradigm. The Agent Based modeling is especially advantageous in the consumer market domain as it allows to leverage the full amount of individual-centric data from the CRM (Customer Relationships Management) systems highly available these days. Although there are no universal straightforward instructions for building Agent Based models, there are certain common steps and patterns. The goal of this paper is to introduce the patterns in consumer market modeling most frequently met in our consulting practice. The modeling language of AnyLogic is used throughout the paper

A hybrid simulation optimization approach for supply chains

The main idea of our approach is to combine discrete-event simulation and exact optimization for supply chain network models. Simulation models are constructed in order to mimic a real system including all necessary stochastic and nonlinear elements. Such simulation models are used as proving grounds for analyzing and improving a real situation on a trial-and-error basis. A traditional optimization method on top of a simulation model has major disadvantages: The optimization method uses the simulation model as a black-box. Information about the structure of the problem is not available and cannot be used for an intelligent optimization strategy

Fully agent based modellings of epidemic spread using AnyLogic

The problem that we are going to conquer subsequently is a slight modification of the ARGESIM Comparison 17. This comparison does ask for the simulation of a SIR-type epidemic by means of lattice gas cellular automata (LGCA). At the end of this paper we will compare the outcome of such an approach with our ABS-result. The task is to model a SIR-type epidemic, an epidemic simplified in several ways. For example we assume a constant population over the whole simulation, thus no births or deaths...

Heterogeneity and network structure in the dynamics of diffusion

When is it better to use agent-based (AB) models, and when should differential equation (DE) models be used? Whereas DE models assume homogeneity and perfect mixing within compartments, AB models can capture heterogeneity across individuals and in the network of interactions among them. AB models relax aggregation assumptions, but entail computational and cognitive costs that may limit sensitivity analysis and model scope. Because resources are limited, the costs and benefits of such disaggregation should guide the choice of models for policy analysis. Using contagious disease as an example, we contrast the dynamics of a stochastic AB model with those of the analogous deterministic compartment DE model. We examine the impact of individual heterogeneity and different network topologies, including fully connected, random, Watts-Strogatz small world, scale-free, and lattice networks. Obviously, deterministic models yield a single trajectory for each parameter set, while stochastic models yield a distribution of outcomes.

How to build a combined agent based / system dynamics model in AnyLogic

  AnyLogic allows you to build a simulation model using multiple methods: System Dynamics, Agent Based and Discrete Event (Process‐centric) modeling. Moreover, you can combine different methods in one model: put agents into an environment whose dynamics is defined in SD style, use process diagrams or SD to define internals of agents, etc, etc. Any kind of mixed architecture is possible due to flexible object‐oriented AnyLogic modeling language. The choice of model architecture (how to partition...

Understanding retail productivity by simulating management practices

The retail sector has been identified as one of the biggest contributors to the productivity gap that persists between the UK, Europe and the USA. It is well documented that measures of UK retail productivity rank lower than those of countries with comparably developed economies. Intuitively, it seems likely that management practices are linked to a company’s productivity and performance. Significant research has been done to investigate the productivity gap and identify problems involved in estimating the size of the gap; for example the comparability of productivity indices, historical influences, general measurement issues, and varying sectoral contributions. Best practice guidelines have been developed and published, but there remains considerable inconsistency and uncertainty regarding how these are implemented and manifested at the level of the work place. Indeed, a recent report on UK productivity asserted that, “... the key to productivity remains what happens inside the firm and this is something of a ‘black box’”. Siebers and colleagues conducted a comprehensive literature review of this research area to assess linkages between management practices and firm level productivity. The authors concluded that management practices are multidimensional constructs that generally do not demonstrate a straightforward relationship with productivity variables. Empirical evidence affirms that management practices must be context specific to be effective, and in turn productivity indices must also reflect a particular organization’s activities on a local level to be a valid indicator of performance