Successful simulation projects have one thing in common. They are delivered by experienced simulation engineers who know the domain and have experience with simulation software technology. There are no shortcuts in simulation, and experience matters.
The simulation process usually starts with someone asking the question: “Should we simulate this process?”. Engaging an experienced simulation practitioner can save a company a lot of time and effort answering the first question. And be careful. The simulation expert who always answers the “should we simulate?” question with “Yes” may not be as experienced as you think. They may have a staff of people they need to keep busy, and see your project as a means to increasing their productivity metrics. Look for someone who will ask important questions about the model objectives, and someone who is willing to tell you that it will not be worth the effort to simulate the system.
Simulation software technology has come a long way in the three decades that commercial simulation tools have been available. But the software tool is not an “easy” button that somehow simplifies building models of complex systems. One of the challenges to model building is choosing the level of model detail. The software does not help one decide which details are important. Watch out for a software vendor that says they can “model any process”.
If you are considering purchasing a simulation tool and building simulation expertise in your company, remember that it may take months of full time work for a person to learn a package and successfully deliver a simulation model. Is your company ready to invest in building simulation capability, and will you have the workload to sustain that person? One rule of thumb I use is that two to three “substantial” simulation projects a year are required to justify building internal simulation expertise. A substantial project would take an expert practitioner one to three months to complete. Yes, I said “months”. There are very few simulation projects that can be delivered in days.
Simulation models can include all of the details of the engineering drawings, historical data, and the control logic used in the system to decide “what next?”. Think of simulation models as a representations of complex systems, combining CAD, spreadsheets, and Warehouse Execution System (WES) control logic into a program that can add random behavior and accelerate time. Let that sink in for a minute. Does that description sound like something that can be made “easy”. Maybe someday, and my bet is that it could happen sometime after we have autonomous cars and people living on Mars.
I would love to hear your comments. Please email me at