Introduction to Simulation Modelling Chapter1

Models as convenient worlds Our lives, as individuals or families or workers in an organization, are increasingly complex. We depend on artificial aids for our survival. We travel by cars, boats or planes; we cook using electrical or gas devices; and we depend on computers on almost all sort of communications. We are part of an interconnected world in which our decisions and those of others can have major consequences for us and for others.

Models as convenient worlds When our decisions turn out well, we and other will benefit. When things go wrong, the results could be disastrous. The same is true for businesses. For example, the cost of a manufacturer’s decision to re-equip the existing factory or build a new one can be huge. How can they be sure the decision will turn out well? How can they be sure the demand for the product will be as expected? And so many questions regarding the technologies used, the location of the factory… etc

Models as convenient worlds The consequences of failures can kill the business. One way to support such decisions is to find ways to learn from failures that could sometimes occur. This learning implies that the investigators have something – a model – against which the performance of the system can be compared.

Models as convenient worlds Think about consequences When we make a decision and take an action, there will be consequences. The consequences may be within our control, or there may be considerable risk or uncertainty. If we have full control then we only have ourselves to blame if things go wrong, and we can take credit when things turn out well. Therefore, it make sense to think through the consequences of our decisions and actions.

Models as convenient worlds Think about consequences The consequences of a simple decision is to list all possible outcomes and choose the most desirable – making sure that we know which course of actions will lead to this outcome. If I want to invest some cash in one of a small number of investments, where the return of each investment is known, simple arithmetic will help me to choose the best investment. This is a simple decision model. (Think of the game of chess!!) Often, life is not that kind and we are at the mercy of other people’s actions or other natural events. So if we make a decision and take an action, our competitors will act as well. Hence, customer’s responses are not wholly predictable.

Models as convenient worlds Think about consequences Does that mean rational and logical analysis is a waste of time?

Models as convenient worlds Think about consequences Does that mean rational and logical analysis is a waste of time? Rational and logical analysis is crucial in our complicated world. However, it might be instructive to think about what other ways there might be to take such decisions, and these include the following:

Models as convenient worlds Seat of the pants: This term usually implies rapid decision-making based on intuition, with no real attempt to think through the consequences. Superstition: This term is used here to indicate a mystical belief that examining some other system will shed light on whatever decision that we are facing, even when there is clearly no link between our decision and the system that we use as a reference. Faith and trust: This term is used to denote an approach that is close to superstition, but with one important difference; that there is some proper link between the system and the decision we are facing. Do nothing: This is the classical “head in the sand” approach, closing our eyes and hoping the problem will go away by any other means.

So, what is a model? A model is a representation of part of reality as seen by the people who wish to use that model to understand, to change, to manage and to control that part of reality.

What is simulation modelling Simulation can be defined as the use of a model to investigate the behaviour of a business system. The performance of the business over an extended time period can be observed quickly and under a number of different scenarios.

What is simulation modelling Given enough time, money, expertise and computer power almost any system can be simulated, however; this may not be sensible. Hence, the first question to face is what type of systems are modern computer simulation methods best suited? The following features tend to characterize the systems best suited to simulation: Dynamic: Their behaviour varies over time Interactive: They consist of a number components which interact with each other. Complicated: There are many objects interacting in the system of interest, and their individual dynamics need careful consideration and analysis.

Why do we use simulation Simulation modelling is used to assist decision-making by providing a tool that allows the current behaviour of a system to be analysed and understood. Simulation can provide the following assistance: Allows prediction Stimulates creativity by allowing different decision options Avoids disruption of the real system Reduces risk of failure Provides performance measures Acts as a communication tool where the dynamics of the system can be visualised over time.

Why do we use simulation Assists acceptance of change Encourage data collection Allows overview of whole process performance rather than local activities Acts as a training tool by demonstrating some of the process behaviour without changing the real system

Simulation and Variability Most business systems contain variability in both the demand on the system (customer arrival) and the duration (customer service time) of activities within the system. Example: A manager of a small shop wishes to predict how long customers wait for service duration during a typical day. He identified two types of customers, who have different amounts of shopping and therefore take different amounts of service time. A: accounts for 70% , and take on average 10 minutes in service B: accounts for 30% , and take on average 5 minutes in service The manager estimated that during an 8 hours day, the shop will serve 40 customers.

Simulation and Variability The service time during theday A: 40 x 0.7 x 10 = 280 minutes B: 40 x 0.3 x 5 = 60 minutes Total service time is 280 + 60 = 340 minutes out of the possible (8 x 60) 480 minutes which counts for 71% Thus the manager is confident all customers will be served promptly using a fixed time between customer arrivals; (480/40) 12 minutes After some calculations on service time: Average: 8.5 minutes Minimum: 5 minutes Maximum: 10 minutes

Simulation and Variability However, in realty customers will not arrive in a fixed time between them as explained. The exponential distribution is often used to mimic the behaviour of customer arrivals. Hence, with a mean of 12 minutes, the calculations on service time will look as follows: Average: 17 minutes Minimum: 5 minutes Maximum: 46 minutes

Applications.. where is simulation Used? Capital Investment Manufacturing Maintenance Transportation and Logistics Customer-Service Systems Business Process Re-engineering (BPR) Initiatives Health systems IT Systems