Agent based modeling focuses on the individual active components of a system. This is in contrast to both the more abstract system dynamics approach, and the process-focused discrete event method.
In agent based simulation, active entities, known as agents, must be identified and their behavior defined. They may be people, households, vehicles, equipment, products, or companies, whatever is relevant to the system. Connections between them are established, environmental variables set, and simulations run. The global dynamics of the system then emerge from the interactions of the many individual behaviors.
AnyLogic combines professional agent based, discrete event, and system dynamics modeling in one platform for efficient, no compromise results.
Agent based modeling is a new way to look at your organization
Traditional modeling approaches treat company employees, customers, products, facilities, and equipment as uniform groups, passive entities, or just resources in a process.
System dynamics models, for example, necessarily contain assumptions such as, “We have 120 employees in R&D, and they can design about 20 new products a year.” or, “We have a fleet of 1200 trucks with a defined monthly shipment capacity, and 5% of them need to be replaced each year.”
Meanwhile, discrete event models view organizations as a number of processes, such as, “A customer calls a call center, the call is first handled by a Type A operator, which takes an average of 2 minutes, then 20% of the calls need to be forwarded.”
These approaches are more powerful than spreadsheet based modeling. They can capture organizational dynamics and nonlinearity, but they ignore the unique composition and complex relationships of individual entities. For example, a customer may consult his family before making a purchase decision, or individual aircraft availability may be determined by rigid fleet maintenance schedules.
The agent based approach is free of these limitations because the focus is directly on individual objects, their behavior, and their interaction. As such, an agent based model is a set of interacting objects that reflect relationships in the real world. The results make agent based modeling a natural step forward in understanding and managing the complexity of today’s business and social systems.
Big data made to work with agent based modeling
Today’s companies and governmental organizations have accumulated large amounts of data in their CRM, ERP, HR, and other databases. Agent based modeling is a powerful way to put that data to work. An agent based model featuring individuals can use real, personalized, properties and behaviors taken directly from these databases. The results deliver refined optimization by providing a precise, easy, and up to date way to model, forecast, and compare scenarios.
Agent based modeling and multimethod modeling
AnyLogic is the only professional software for building industrial strength agent based models. Moreover, agent based models can be easily combined with discrete event or system dynamics elements, for complete, no compromise, modeling. This can be seen, for instance, with warehouses which behave on a supply chain as agents, but are modeled internally using discrete event modeling.