Because of the rapid growth and increased competition in information technology (IT), business process outsourcing (BPO) and other service sector industries in India, quality and cost of operations have become the major distinguishing factors among such companies. Survival, growth and profits depend on how an organization controls its costs and satisfies its clients or customers.

Many organizations have adopted quality improvement programs, the important ones being Six Sigma and Kaizen. They also have modified the techniques of these programs to best suit the organization’s needs. To generalize, the choice of the quality philosophy has been made on such factors as scope and duration of the projects, the organization’s product or processes, and the statistical intensity required to analyze and improve.

Irrespective of the quality program used, many organizations have found limitations in some of the quality improvement tools they use. At the same time, they are discovering the advantages of using simulation modeling and analysis as a problem-solving tool.

Limitations of Quality Tools Used

The reasons why companies are finding that some analysis methodologies provide sub-optimal results include:

• Complexity of the System Under Study – The business scenario has become highly complex with continuous changes with which organizations must cope. When initiating a project on quality for a highly complex new or existing system, often there are too may factors affecting the performance of the system. Even Six Sigma may fail as it becomes impossible to statistically analyze the system or provide statistical alternatives to the existing system. This has prompted project teams to provide ad hoc alternatives as solutions.
• Sensitivity or Robustness Required – Analysis methods provide a solution to the problem at hand, but a slight change in input or a minor business decision requires the quality project team to “reinvent the wheel” by kicking off a new project to solve the “new problem.”
• Verification of Analytical Solution – There is no pedagogical pattern to reinforce or verify the solution arrived at. Most quality methodologies include having to implement and measure the solution provided to determine if the required quality level (or Sigma level) is reached and then control the system to stay at that level. If the project has not met the expectations, it will need to be restarted. This is highly costly to the organization which must change the processes or work force or even make business decisions based on the project’s analysis. Cost also is incurred when actual experimentation (design of experiments) is done on the system.
• Inability to Analyze a Stochastic System – When the outcome of an activity can be described completely in terms of the input, the activity is deterministic. When the effects of the activity vary randomly over various possible outcomes, regardless of the complexity of the system, the activity is stochastic. Many systems currently used in the industry are stochastic and cannot be easily modeled or studied in the current quality methodologies. The solutions provided to such systems are ad hoc and never satisfactory. Statistical modeling is necessary to study such systems.
• Inability to Visualize the System – When studying a system for bottlenecks, lead-time reduction and process changes, it can become difficult to visualize it. The quality team requires a scale model to assist it in spotting bottlenecks. Mere numbers such as average handling time (mean time) or standard deviation can be misleading. Even system changes need to be visualized.

These limitations of quality processes can be dealt with by implementing an operations research technique called simulation modeling and analysis.

An Introduction to Simulation

Simulation is imitation of the operations of a real world process or system over time. It involves the generation of artificial history of the system and the observation of that artificial history to draw inferences concerning the operating characteristics of the real system.

Operations research and simulation modeling have been used in the past by upper management for decision-making in various areas, including supply chain management, manufacturing applications, semiconductor manufacturing, construction engineering and military applications.

Simulation is now in use in service industries to model and analyze call flows, human resource management and forecasting. Usage had generally been a one-time effort due to various disadvantages of the simulation concept, but current technology and development have actually converted these disadvantages into advantages. Some of the important ones are:

• Data Availability – Simulation requires a large amount of data. In the past, data was generally not available and had to be collected, which was a strenuous activity and took a lot of time. Now the use of enterprise resource planning software and customer relationship management programs provide large volumes of data. That data can be used as input for simulation.
• Cost of Modeling – In the past, companies developed their own simulation software to supplement their analysis. That software was costly to procure. Now, many off-the-shelf simulation software products have been developed. They are cheaper and easy to use, and can be applied in different business scenarios. The software packages also provide graphical representations of the model.
• Extensive Knowledge of Probability and Statistics – Simulation modeling requires the use of probability and statistics to model the system. This was a hindrance in the past as many system modelers were reluctant to bury themselves with probability and statistics. The current simulation software packages have input analyzers – this also is available with many quality software tools – which help with the data conversions. Only a working knowledge of statistics is a prerequisite to effective simulation modeling.
• Time to Run Simulation – Thanks to the current computer processor speeds, simulation can be run quickly or even be slowed down to assist the project team.

Integrating Simulation Modeling and Analysis

Most organizations have modified the quality techniques to suit their requirements, but the basic project methodology for continuous quality improvement remains. The projects would follow the basic outline as shown in Figure 1.

Simulation modeling and analysis as a tool can be best used in Steps 2 and 3 in the framework above. It is most useful when studying the system, designing the system, evaluating alternatives and backing up the results of the improved process.

A typical example of how it can be done is shown in Figure 2 for a Six Sigma DMAIC (Define, Measure, Analyze, Improve, Control) process. DMAIC applies to an existing process that needs improvement. It is best applicable to continuous defect reduction in a cross-functional/uni-functional environment.

Conclusion: Enhancing Current Methods

Simulation modeling and analysis can be used in a quality improvement framework as an enhancement to current methods. Some key points to remember when deciding to use simulation are:

• GIGO (garbage in, garbage out) applies to simulation. The way the system is modeled and data is entered will determine the efficiency of the model itself. This will mean that the project team or at least an analyst must be trained in the use of simulation. The analyst or team must know the system well enough to gauge the factors and level of detail to be simulated.
• Simulation must not be performed if the team does not have time or resources for a detailed quality project. This would obviously mean a sub-optimal solution or a stopgap solution.
• Various off-the-shelf simulation software is available in the market. A detailed feature study needs to be made before purchase so the software will best fit the organization’s needs.
• The simulation software is still a cost to the company. The monetary gains will be seen only after successful completion of the project.

Simulation modeling as a tool currently is the best addition for a continuous improvement process. Organizations have a lot of new challenges when it comes to quality of service. These new challenges can only be dealt with by taking it up with newer ways of finding solutions. The quality framework needs to be upgraded as the situation demands.