I am a simulation modeler, that strives to build useful and realistic models based on facts – how things “really work”. I spend as much time as it takes digging through resources like operation manuals and policy guides. More importantly, I spend time with the people who know how decisions are made using what information. Gathering up all of the bits required to build a useful simulation model requires serious skills and requires serious time.
What is really bothering me is how to best capture and organize all of this critical information so that it is organized and available? The information that I collect comes in many forms including published documents, annotations, primary source quotes, flow charts, spreadsheet tables, and more. While some of the information is digital much of it is in the form of notes, drawings, or even recordings. Complicating the task is keeping track of what was derived from what (supporting conclusions or narratives); or supporting the validity of viewpoints using multiple independent sources.
Of course, I have dealt with these issues for awhile now. I keep careful records filed in carefully named folders that lend themselves to locating my sources. Most of my folders are digital. In addition, many notes immediately become part of the model structure, allowing me to put the note aside. If I need to refer back to a source note, I can generally find it and others like it in my files. This method has served me well up to a point. However, the shortcomings have become more and more apparent to me through the years.
A big issue is a project delay. A lengthy break in model building allows my recall of familiar information to weaken. I find myself reading back through files to refamiliarize myself, but often important context has been lost. I have struggled with a memory of something important being said, but being unable to quickly put my hands on it. Lack of familiarity is a problem immediately if a new teammate joins the effort, or if a large modeling effort is divided into parts.
Another issue is the flatness of organizing in folders. Each piece of information is categorized in only one way: the name of the folder. This is like a checkbook register. If I wanted to know what I have spent on groceries I go back through the register and add it up. If I want to know the amount of a certain deposit, I need to remember when the deposit was made or search the whole register. This is a limited and frustrating search technique that is out of date. Using personal financial software, I can search my check register by date, amount, payee, category, etc… So, why can’t I organize my critical modeling information the same way?
Well, after a couple of years of researching, thinking, designing, and building (in my spare time), I had created DynamicVu: Information Management for Simulation Modelers (just me at the time). DynamicVu is a secured on-line database designed for individuals and teams that build useful simulation models built on facts. Instantly recalling support information creates faster model development and stronger model confidence.