DAMAS: An Integrated Business Modelling Information System to Support Management Enterprose Decisions
چکیده
DAMAS aims at building a decision support system based on the integration of legacy data repositories and system dynamics modelling tools. The latter are used to simulate the behaviour of different business areas, such as marketing, finance, production, etc. Moreover, an enterprise-wide, company-specific model, integrating the aforementioned areas, is being built. The main target is wine production industry. Nevertheless, the DAMAS consortium is pursuing the applicability of the proposed approach to other industrial sectors. DAMAS features a business object architecture, encapsulating legacy repositories, as well as business intelligence and common functions. Managers are provided with a high-level dashboard which co-ordinates, controls and monitors the underlying business objects. 1. GENERIC PROBLEMS IN DECISION SUPPORT SYSTEMS Still today the majority of software systems supporting every-day work in business company, are conceptually of “metabolic” type, i.e., these systems are procedural-minded: they use a huge amount of data to produce, mainly through aggregation procedures, documents (e.g., invoices and orders) rather than information. Even when a company information system of this type is well designed, top managers do not receive much information: usually aggregated data and statistical analysis derived as by-products of the main information management activity. Moreover, each of these sets of data typically concerns one specific area of company, i.e.; no view concerning the whole enterprise is provided. If in the past years it was commonly accepted that the strategic decision making process of a company could be based on a “sum of elements” view, today the increasing competitiveness of the world-wide market is demanding for a Systemic Approach. The corporate management, which has the aim of controlling a complex system (made of relatively autonomous, though continuously interacting sub-systems), cannot be effectively supported by the views of single sub-systems. In fact, even the management of a specific department should analyse the effects of its decisions over the whole company, since optimising the behaviour of single departments does not generally result in maximising the efficiency of the whole organisation. In particular, two essential concept time (dynamic vision) and feedback reactions must be taken into account. Considering these factors allows the decision-maker to evaluate the potential consequences of a decision, in both the areas of competence and in all other business areas including external environment. 2. DECISION PROBLEMS IN WINE INDUSTRY Competition is constantly growing in the wine production industry, especially because of the increase of high-quality producers: traditional European wine producers have been recently joined by producers from North America (particularly California), Australia, New Zealand, South Africa and South American countries. As a result of this situation, wine companies are looking not only for innovative production processes, but also for effective and reliable management techniques. However, increasing efficiency in the wine industry is not easy, because of the relative complexity of wine industry processes. For example: • Vintages vary in quality and quantity; crop features depend on the weather and on factors not entirely understood. • The optimal allocation of crop from different vineyards to different factories is a complex function of capacities and maturation rates. • Wine production involves lengthy delays, forcing the winemaker to decide on production volumes approximately 18 months before the product is released to the market. • Demand for a vintage is very difficult to judge, being dependent on consumer taste and market perception. As a result of this complexity, managers need support tools to understand the influences on their business and the long-term effects of their decisions. System Dynamics models is a powerful tool supporting managers in this task. However, given the non-technical background of most managers in this industry, part of the challenge is to make the required technology accessible to those who must use it. 3. THE ROLE OF SYSTEM DYNAMICS The System Dynamics methodology was originally designed to model the behaviour of complex systems over time. It does so by representing the processes, structure, strategies and information flows. A definition (Wolstenholme 1990) of the method can be stated as follows: “A rigorous method for qualitative description, exploration and analysis of complex systems in term of their processes, information, organisational boundaries and strategies; which facilitates quantitative simulation modelling and analysis for the design of system structure and control.” System Dynamics (SD) is a methodology that enables decision-makers to better understand problems as well as the structure of systems to be managed. In fact, it makes possible the identification of a limited number of master variables (company levers) on which decision-makers have to focus their attention. Moreover, it helps to reach a homogeneous level of aggregation among variables belonging to different functional models, thus achieving an integrated view of problems and, consequently, an inter-functional perspective embracing different business areas. 3.1 What is System Dynamics System Dynamics models are actually systems of differential equations, such as: L(t) = L(t-∆t) + (Rin (t-∆t) Rout(t-∆t)) ∆t L(t0) = K Rin (t) = f(Rx(t), ... Ry(t)) Usually it is not possible to find analytical solutions, because equations that describe real systems are generally non-linear. However, it is possible to simulate the behaviour of a system described by SD equations, starting from the known initial status (e.g., in the equations above the value of L at the initial time t0 is K) and computing the value of variables at time t0+∆t, then at t0+2∆t, etc. Provided that equations represent correctly the nature of the system, and that ∆t is chosen small enough to capture the dynamics of the system, the simulation actually represents the behaviour of the modelled system. By the way, there are several commercial computer programs that support SD model editing and simulation. Of course, models generally represent dependencies from "external" variables, whose values in the whole simulation period are supposed known in advance. These variables typically represent "scenarios" representing phenomena not depending on the system being modelled. For instance, the arrival of grapes at a wine production site depends mainly on weather. Decisions concerning the capacity of the production plant should thus be based on the simulation of several possible arrival scenarios. 3.2 System Dynamics in DAMAS In DAMAS the role of System Dynamics modelling is crucial. Each business area is analysed according to SD methodology, and the feedback influences with main variables of other areas are explicitly represented. The architecture of different business area models (namely Production, Marketing and Accounting) is presented in the Figure 1. Two different level models are developed for each area: a detailed model which takes into account all the interactions among variables of interest to the specific business area, and an aggregate (or macro) model taking into account variables which mainly influence other areas. This architecture preserves the detailed view at the business area level without loosing the overall company systemic scenario. Production Business Area Model Marketing Business Area Model Accounting Business Area Model Production Model Macro Marketing Model Macro Accounting Model