SDM ’07 Certificate Program
Editor’s note: Rupesh Goel is a student in SDM’s one-year, career-compatible graduate certificate program. The program is designed to provide companies with a rapid infusion of systems thinking and to update the toolset of senior engineers. Certificate program students typically attend classes at a distance—Goel is participating from India—and work on a capstone project of interest to their employers. In this article, Goel describes his SDM project.
I work as practice head for simulations at Tata Interactive Systems (TIS), a company that provides simulation software for business process optimization as well as risk/change management training. Last year my team was assigned an interesting project—we were to produce a new technical architecture for an older product platform (TOPSIM) that the company had recently acquired.
Our goals were:
• Re-architect the application layers to allow for customization in the client front end
• Infuse new technology into the existing product
• Increase reusability of the code source for future application development
We also wanted to improve the platform’s usability andoverall features. Essentially we set out to design a second-generation TOPSIM without starting on a clean sheet.
A typical simulation development project requires the teamwork of software engineers, a system modeler,
graphic and instruction designers as well as usability experts. Programs must be delivered on time, on budget,
and with minimal risk. Reusability (from past projects and for future projects) is a key metric apart from the operation project metrics. I had both management and design roles on the team.
TOPSIM is a business simulation platform with complex parts, deployed under various scenarios, including in workshops and on the web. It really is a system, with several moving parts—including an authoring platform for creating new what-if scenarios, a user system for playing the simulation, as well as modules that help optimize, debrief, and report. The simulation can be made competitive and multiplayer to simulate economic and market models.
The delivery of the product also relies on support of personnel, printed material, web-based learning assets, and network equipment to support multiplay. At the heart of every TOPSIM is a systems model of the business process and entity that is customizable.
First we benchmarked TOPSIM against other vendors in the market. SDM Professor Edward F. Crawley’s framework for critiquing and re-architecting system architecture was really useful in arriving at the basic set of changes, viewed from a user perspective. We also ran a survey to capture both software delights and pains, in addition to overall user experience. We used quality function deployment to capture our first-cut analysis.
We then crafted new system performance goals and, through a variety of design tools (a design structure matrix was actively used), we arrived at a new design. We also evolved new process maps for the modeling, design, and development of the new architecture. Care was taken to allow for the various “ilities” as they are key to the revenue stream of software applications.
The new architecture also required us to approach development, especially of the model for complex numeric model, in a different way. This change is an ongoing one, but we have already defined new templates for the modeling which allow us to develop a prototype in Microsoft Excel and import it into the new architecture. We are now investigating a change to a visual way of modeling, customized to our needs, in which modelers can show the flow and design of a complete business process through a graphical interface. The model could then be imported into an object-oriented source code library.
The new product architecture is now being tested with one of our clients in Florida.
As we discussed the project in the MIT SDM certificate reviews, we also realized that the visual modeling method when combined with an Applications Service Provider delivery model could be used to generate a totally different revenue stream—from Internet-based users who can access the solution from different locations, which allows for scalability as well as authorship by independent consultants. MIT faculty and reviewers gave invaluable input on how this could be commercialized, and my group is now preparing the first set of specifications on how to take this to the Internet-based model.