Guidelines for Selecting a Simulation Engineer
As the use of process flow simulation increases, organizations are trying to find individuals to be Simulation Engineers. In this post, we discuss the desired capabilities and skills/experience to look for in a simulation engineer. Let’s start with the most critical desired capabilities for a simulation engineer:
- Understand which critical components of a process should be modeled: The initial challenge with developing a simulation model is determining what operational details to include. For simulation neophytes, the response is to include as much detail as possible. Unluckily, the more details you include in the model, the longer it takes to develop the model, and the longer it takes to run experiments with the model. A good simulation engineer knows how to understand enough about the process to identify the critical components of the process that should be modeled.
- Structure the model code in a manner that facilitates continual use/expansion of model: Too many simulation models are one use only. A good simulation model should be used first to identify failure points in a process, and how to avoid the occurrence of these failures. Then, it should be used to evaluate operational strategies including scheduling logic.
- Know how to select distributions for processing times: A big benefit of using simulation is the ability to reflect the random behavior inherent in our processes, and this requires that we utilize statistical distributions (vs. a specific numeric value) for most of the model’s data. Unluckily, most organizations do not collect accurate process time data. A good simulation engineer should be able to manage data collection efforts, fit statistical distributions to data collected, and select an appropriate distribution when sufficient data does not exist.
- Interpret the results of the experimentation: The simulation engineer needs to accurately select the model’s run-time parameters (number of replications, length of warm-up period, and length of model run), design efficient sets of experiments to run, and interpret the results using confidence intervals.
- Guide the simulation project team: The simulation engineer needs to clearly understand the model customer’s problem solving needs, guide the simulation project team to select the appropriate level of detail to be modeled, and coach the model users on how to most effectively experiment with the model & interpret the results.
To get a simulation engineer with these desired capabilities, the skills/experience to look for are:
- Industrial Engineering degree/background
- Probability and Statistics
- Programming skills
- Communication/coaching skills
If you have additional perspective to share, please do. Thanks, Khaled
Very nice description of the ideal candidate pedigree! The only thing I could add to this: soft skills. Being able to capture Client needs, educate a Client that don't know about simulation and communicate to non-simulator people. These aspects can make a huge difference when regarding simulation's credibility.
Matthew touches upon a very important point here: simulation projects handed off to an engineer who recently joined the team because he knows his/her way around a computer. In reality though, managing your customer is equally important as managing the project itself. The art of simulation is to take your customer on a journey of process improvement of which simulation is the tool of choice. By guiding a customer through the simulation project you'll uncover so much new data and so many interesting new aspects of the system that all significantly contributes to the value added by the work. An experienced professional will be able to see all of these inherent benefits long before the emerge for the customer and it is important to guide the customer through this process as well as making sure they focus on their actual needs. Too many times customers can get carried away with the potential of simulation that they need to be reigned in to ensure an on-time delivery of the project. Any additional work can easily be done after completion of the original work. The experienced professional will be able to discuss this with the customer and pursued him/her to follow a focus path towards a successful project.
Benjamin - I agree that most young people that may be interested in STEM careers don't know that simulation exists. When I am talking to high school or college students, I usually open with "I play computer games for a living." It definitely gets young people's attention! So, step (1) is getting them interested, and I think step (2) is getting them prepared. I like the idea of an "apprentice network" that Matt mentions. I think encouraging the full time, professional simulation consultants to be project sponsors or provide guidance for senior design projects could really help us see more students prepared for a career in simulation. Matt, since you are involved in WinterSim, maybe this is something we can have on the agenda as a round table or panel discussion for next year? :)
Thanks Khaled, nice summary of what makes a simulation engineer. Sadly too many companies put more energy into the selection of a simulation tool than they do into the people who will be doing the work. How many times have we seen the project handed over to the new engineer who knows his way around the computer? Simulation thought leaders like yourself need to continue to re-frame the focus on the engineer. And I believe we need a "training academy" and/or "apprentice network" for simulation professionals.
Agree with your 4 capabilities. It is a hard mix to come by - detailed & big picture - ability to program & give a killer presentation. What can we, as an industry, do to grow talent in simulation modeling? I have seen too many companies fail with simulation modeling because they think anyone can do it internally or the cheapest/faster consultant must be the best for them.