warehousing simulation modeling
Distribution and Warehousing

Business Success with Simulation

Ensuring Success with Simulation

Simulation is an approximation of reality, the level to which it mimics reality depends on the tool and method used to implement the simulation model. The closer the simulation model mimics reality, the better its results and analysis are, while the faster the model is built and validated the greater the return on investment.

Introduction

Successful simulation projects have had significant return and savings that far exceed the cost of the project. Moreover, the benefits achieved are normally long term and continue to provide savings for years to come. Unfortunately, not all simulation projects are successful and many companies have had their share of simulation failures and lost investment.

This white paper lays out a set of rules and procedures to help achieve constant success with simulation and to ensure that the benefits achieved last for many years. The procedure is simple, yet when followed, will achieve amazing results no matter how complex the problem is.

Setting A Goal

The key to success for any project is the clear definition of the end result. Whether it is a complex software application, a construction project, or a process improvement implementation having a clear definition of the project goal is key to its success. Throughout every project, a constant reminder of the goal is attached to every milestone. Improvements and changes are implemented through change requests and are only implemented if they do not derail or fit the initial project goal. The same concept applies to simulation projects.

Simulation in general is graphical while Dynamic Simulation is very animated providing a clear view of the problems, solutions, and everything in between.

Unfortunately, the clear visibility provides a venue for people to look beyond the initial problem and start questioning other aspects of the operation that are not part of the initial goal. Those aspects, although related to the overall operation, should be left aside until the initial goal is attained except where they have a direct impact on the current goal. Therefore, instead of modifying the current goal and expanding it midway through the project, the project manager should add and prioritize the additional issues as additional milestones following the completion of the initial target goal.

Involving The Customer

Along with setting the goal and staying the course, it is imperative to constantly communicate the findings and potential solutions with the parties affected by the change. Buy-in from the affected stakeholders must be achieved in order to fully realize the benefits of the proposed solution and maximize the return-on-investment. Communication can be achieved through weekly summary sessions and an explanation of the status of the simulation and issues that have come up.

Another key aspect is listening to and addressing the concerns that are brought up by the customer or the group being affected by the change. Failing to do so will alienate the group and instead of gaining buy-in from the team, it generates resentment and resistance to change and improvement.

Data Availability

The availability of accurate data is critical to any simulation project but not all data is required to ensure its success. A key understanding of the process flow, object interactions, and resource/manpower constraints are all important aspects to every simulation model. In addition, cycle times, transfer times, as well as model and process constraints are critical in ensuring the model behaves correctly and is properly validated. A mistake that is often made is that wait times are entered into the model when in fact they should be a validation tool to see if the model behaves similarly to the live environment.

Exceptions need to be studied carefully before adding them to the model. An event that occurs once in 12 weeks should not be modeled if the simulation run only covers one week. Instead, the event should be studied as a special case scenario and the data should not be a part of the main model.

Model Building and Validation

Model building and model validation should be performed in a cyclic process. As new constraints are added to the model they should be individually and immediately validated. Dynamic simulation tools excel in this environment due to their ability to allow the user to build and interact with the model as the simulation is running, hence visualizing and validating the change as they occur.

The model building cycle needs to be controlled and targeted to the initial goal of the model. Adding additional constraints and behavior that does not impact the goal should be avoided and deferred to a later stage in the analysis process.

Overall Model Validation

With the model ready, a complete model validation is needed in order to make sure that the model behavior and output is in line with the current environment. A sample input sequence should be retrieved from the current environment and run on the model. The resulting data should be within 1 or 2% of the actual data collected from the line.

Scenarios, What-Ifs, Defining A Solution

With the model built and validated, scenario and what-if analysis should start. If the tool being used does not include a scenario analysis environment, the user should export the data to external tools for tabulating and comparison.

After a scenario is selected as a potential solution a number of steps must be taken to ensure that the elected solution will perform as expected.

Previous Article Modeling Simulation Software
Next Article What is Simulation?
Print
13341 Rate this article:
5.0
Comments are only visible to subscribers.