RESEARCH & RESOURCES

Use Simulation to Make an Impact

To explore future possibilities and opportunities that do not currently exist, enterprises must explore the capabilities of simulation.

By Mark Peco, CBIP

[Editor's Note: Mark Peco is leading a session on simulation at the TDWI World Conference in Orlando, December 8-13, 2013. Simulation Methods for Business Analytics: Techniques for Driving Impact and Value introduces computer simulation techniques as well as the basic concepts and a framework that positions simulation techniques within a broader BI program.]

Introduction

The traditional concept of business intelligence (BI) was based on the principle of gathering data that already existed in an organization and putting it into a suitable structure that allowed business managers and staff to access and format it into meaningful reports that answered many of their questions about business performance.

This principle is constrained to only using existing data to answer as many relevant business questions as possible. However, many valuable questions cannot be answered from existing data sets alone. The existing data describes the status quo of business operations. It represents how existing processes are performing currently and historically. Existing data cannot describe conditions that haven't existed. In other words, if the scope of a BI program is constrained to include only existing data produced by the organization, how can new opportunities be identified? How can new ideas be explored? How can improvements to existing process and performance be identified?

When an organization has the maturity to explore how to improve existing processes or to identify and implement new processes, how can they proceed by analyzing existing data? To analyse conditions that haven't actually occurred, additional sources of data and analytical capabilities are required. The need to explore future possibilities and opportunities that do not currently exist can be addressed by the capabilities of simulation.

Simulation is not a new concept. As a discipline, computer-based simulation techniques have been used for several decades. Simulation has routinely be implemented to design new facilities, decide how to allocate resources, predict future operations, and help managers deal with abnormal or unexpected operations. Historically, simulation models have been deployed and operated in organizations in an ad hoc manner by operating groups, planning groups, design groups, and marketing groups. It has typically been implemented within local operating pockets of an organization to solve and address functional problems by specific groups and departments.

The Convergence of Business Intelligence, Analytics, and Simulation

However, a new concept is rapidly emerging. It is based on the convergence of business intelligence concepts with the existing and expanding base of simulation capabilities that have existed in pockets for many years.

This emerging concept understands that if simulation and business intelligence become better integrated, the BI footprint becomes more valuable, and simulation capabilities increase. Synergies will develop as the BI fields of data, information, and analytics converge with existing simulation capabilities under the governance of a well-managed program. Simulation capabilities will grow across their current organizational pockets and become enabled by data management capabilities from the BI programs. As simulation capabilities evolve within the BI program, additional insights can be generated which will lead to high business impact and value generation.

Simulation is an analytical process carried out by people using software-based models also developed by people. Simulation enables experimenting, exploring, and examining new ideas, all within a virtual environment. The virtual environment is enabled by implementing different types of analytical models in software. The models relate input variables to outcome variables based on known rules, partially understood rules, and heuristics. The rules define relationships between groups of input, intermediate, and output variables. When the models are implemented in software, the environment allows people to enter different combinations of input variables to explore how the desired value of output variables can be achieved. The rules of behavior can span time periods and consider degrees of uncertainty using statistical concepts.

The simulation environment becomes a virtual laboratory that enables business experts to explore options, consider new strategies, compare scenarios, and drive decisions that result in desired outcomes.

The Simulation Team

Simulation capability is developed and enabled within an organization by a virtual team that collectively maintains the appropriate business knowledge, technical skills, and modeling experience. Simulation environments include core analytical models and the software platform necessary to exercise the models. They are created by team members with skills in model building, data management, and software development. However, the ability to use the environment to answer difficult business questions and generate meaningful insights requires other team members who understand the business domain, company policy, process details, problem solving, and how to use the simulation models to produce meaningful results. A high-performing simulation team requires roles and skills in the following areas.

Simulation Model Building and Application Development Team

People in the following roles provide expertise and skills needed to build a simulation environment.

  • Domain expert: A business analyst, engineer, operations planner, or decision maker with expertise and knowledge in the functional and process areas that are within the scope of the simulation model.

  • Data analyst: A business analyst, data steward, or IT professional having data management, data quality, and metadata management skills. Knowledge about the meaning and location of the relevant data sources is also essential.

  • Model developer: A combination analyst and modeler skilled in the modeling techniques selected for the initiative. Techniques may require knowledge of physics, engineering, mathematics, statistics, probability, data mining, system dynamics, business processes, queuing theory, and other relevant modeling approaches.

  • Software developer: A developer who can create software applications, databases, and reporting solutions to implement the simulation models into a robust decision-support platform.

  • Model maintainer: Analytics staff with skill in model calibration, testing, and evaluation.

Simulation Usage, Analysis, and Decision-Making Teams

The following roles contribute the expertise and skills needed to harness the power of the simulation environment and translate it into true business impact.

  • Simulation analyst: A business planner or operations analyst skilled in designing simulation experiments, executing simulation runs, and analyzing the output of simulations.

  • Decision maker: A business executive, manager, or operation staff member skilled in evaluating options and making decisions.

  • Operations analysts: A process improvement specialist or knowledge worker with expertise and knowledge about the activities and processes of the organization.

The Simulation Environment

A simulation environment is data intensive. Simulation capabilities require data governance approaches to ensure that the input data used to produce the core models meets quality and integration requirements. Simulation experiments generate large volumes of new data and information that must be managed to support downstream analytics, reporting, and story-telling.

Simulation depends on successful data governance, integration and quality management provided by a BI program that depends on simulation to drive innovation, create new business ideas, and enable high-quality decision making. Making an impact on business and moving the needle meaningfully within an organization depends on many of the analytical capabilities enabled by simulation.

Use simulation techniques to move the needle in your organization. It is essential that the business, technical, and functional reasons for undertaking simulation are fully understood to ensure that the correct form of modeling is employed and that appropriate assumptions are made as models are developed.

There are many opportunities for moving the needle that are enabled by simulation techniques. You can:

  • Identify and evaluate the components of a business strategy
  • Frame, analyze, and diagnose root causes contributing to a business problem
  • Determine target values for defined business goals
  • Experiment to learn and discover new causal relationships
  • Determine how to remove the bottleneck in a business or industrial process
  • Decide how to best allocate resources
  • Create virtual measurements
  • Forecast future operations of a new production plant before it is constructed
  • Identify and evaluate different design options for a new facility
  • Plan and analyze how to accomplish specific objectives
  • Implement process monitoring and surveillance capabilities
  • Enable organizational learning opportunities

The major business reasons for using simulation techniques include planning, monitoring, forecasting, prediction, measuring, optimizing, diagnosing, and learning. Each reason defines an area where the needle can be moved using simulation to drive meaningful impact.

Ensure your BI program is making maximum impact. Implement simulation capabilities as part of your managed BI ecosystem.

Mark Peco, CBIP is an experienced consultant, educator, manager and analyst. He holds graduate and undergraduate degrees in engineering from the University of Waterloo and has led numerous consulting and software development projects helping clients to adapt to fundamental shifts in business models and requirements. His experience includes strategy development, business intelligence, data warehousing, compliance, analytics, application development, simulation, operations and program management.

As a leading practitioner in the field of business intelligence, Mark is on the faculty of the Data Warehousing Institute (TDWI) and delivers courses and workshops to major organizations on a global basis. Mark's professional focus is at the intersection of business, operations, and technology within the context of the energy industry.

Located in Toronto, Mark is an independent consultant and can be contacted at mark.peco@gmail.com.

TDWI Membership

Get immediate access to training discounts, video library, BI Teams, Skills, Budget Report, and more

Individual, Student, & Team memberships available.