The discipline that studies business processes and optimization is often called ‘Operation Research’. This is a general term that implies a number of different techniques. System dynamics is one of the most commonly used frameworks for modeling business processes including the plant operation and the market behavior. However, the system dynamics can only deal with continuous processes and cannot capture the system details. This is a high-level method mostly suitable for studying macroscopic level processes. A list of software for the system dynamics modeling includes:

  • VenSim
  • PowerSim
  • iThink
  • AnyLogic

The field of the business process modeling grew up with deployment of discrete element methods. The discrete element methods were known before for decades and, probably, pioneered by Carl Adam Petri with his Petri Networks. The wide adoption of the discrete element methods was limited with a computer software availability. After the software became available (see the list below) the application of the discrete event modeling, also known as simulation modeling, boomed. The software for the discrete event modeling:

  • Arena
  • ExtendSim
  • SimProcess
  • AutoMod
  • Enterprise Dynamics
  • FlexSim
  • AnyLogic

AnyLogic software distinctly stands out of this list since it combines both methods (system dynamics and discrete events) plus includes integrated agent-based modeling. All modules in AnyLogic can be extended with a user code in Java which, basically, remove any limits for the modeler.

To my opinion, discrete element modeling is a ‘must have’ for all plants and warehouses. Most of Fortune 500 companies use it to test scenarios, optimize work schedules, optimize inventory allocation, and simply monitor the process. A discrete element model of a plant is a full-scale replica of the plant with all details captured: movement of parts on conveyers, assembly lines with workers, warehouses with forklift trucks etc. The model is literally a ‘virtual’ plant on your desktop. All the model elements can interact and operate as it happens at the real plant using realistic work schedules, maintenance schedules, stochastic breaks, etc. The degree of the model part refinement is almost infinite. After the level of details in the model is enough to capture real processes and match real key performance indicators, the model can be used as a decision support tool. For example, a manager can learn what disruptions may occur in the workflow if the plant will experience certain delays with a spare part supply, what bottlenecks will appear in the process if the production output will be increased in two times, etc. The model operation does not require any knowledge in the discrete event modeling. As soon as the model development is done, the model can be saved as a standalone application with user friendly controls.