Scientists from the Universidad Carlos III de Madrid (UC3M) have developed software for a business simulator that emulates how a market behaves with regard to the decisions that company directors make, with applications for education as well as the working world.
SIMBA (SIMulator for Business Administration) has been developed by the UC3M Automatic Planning and Learning Research Group in close collaboration with Simuladores Empresariales S. L., a spin-off technology company created by the INNOVATIC Research Group, from the Department of Business Organization of the Universidad Autónoma de Madrid (UAM). Its developers are a multidisciplinary team of professionals, professors, and information technology engineers, with more than 20 years of experience developing these kinds of tools.
This simulator emulates the reality of companies using the same variables, relationships, and events found in the business world. “It can be utilized from a competitive point of view since different companies can compete against each other to improve their own results”, Fernando Fernández, one of its creators from the UC3M Information Technology Department explained. “The simulator”, he continued, “covers the principal functional areas of companies and allows for different levels of difficulty, depending on the users, who have a very high level of interaction with the simulator”. This interaction is carried out through Internet, and thus the simulator can be used in different pedagogical scenarios, ranging from traditional classroom learning in computer rooms to distance learning courses.
There is no limit to the number of companies that this tool is able to simulate, and it allows different products, markets, client segments and technology to be defined. SIMBA includes other relevant characteristics, such as the capacity to personalize for different institutions, the use of different languages, currency, or socio-economic environments. In addition, it includes a large variety of graphs and data to show principal indicators for companies and markets. All of the foregoing makes SIMBA an excellent tool for the students who are already using it to understand the interactions that are produced between the different variables that appear in the business world. “They can manage their companies successfully or lead them to ruin according to the appropriateness of their decisions in the business environment defined in the simulator. Thus, students can learn from their own mistakes and successes”, Professor Fernando Fernández noted.
Artificial Intelligence applied to a Business
The companies involved in a simulation can be controlled by human beings or by computer programs. These computer programs are developed using Artificial Intelligence techniques such as Automatic Learning, an area which allows computer programs (Virtual Agents) to be constructed, which improve with experience. In this case, these Virtual Agents learn to run the companies in a way similar to the human students, that is, by trial and error. “This increases the complexity of the simulated market and creates the possibility of carrying out multi-player simulations, where some players are human, but where others are controlled by computer programs built by Artificial Intelligence techniques”, Professor Fernández pointed out.
In addition to Automatic Learning techniques that the scientists are using, there are many other tools within the field of Artificial Intelligence which can be used, such as Business Intelligence. The simulator implements a company model which includes basic rules, such as if the price of a product increases, the demand for the product tends to decrease, and as such, less of it will be sold. Nevertheless, the business model includes other very complex elements which also must be codified. For example, in order to develop the simulator, that model is formalized by experts in business administration through the rules of business, and is transmitted to the program developers. The process for the formalization of the model, especially updating and maintaining it, is very costly, given that every time one wishes to make a change in the business model, the code or its implementation in the simulator has to be changed. Currently the researchers are evaluating the use of production systems or systems based on rules, one of the most studied Artificial Intelligence techniques, which allow all of this knowledge to be formalized and implemented by business administration experts in a simple and more straight-forward fashion.