Inhaltszusammenfassung

Simulation researchers have to be mathematicians but also motivators in one person: for the acquisition of real-world projects, they must handle assumptions concerning simulation like being abstract, costly, too complicated, and disconnected from operational needs. And there is some truth to these assessments. You need to build models, collect data, conduct the simulation, and afterwards you need to put the results into operation by implementing software and sometimes hardware. All this costs...Simulation researchers have to be mathematicians but also motivators in one person: for the acquisition of real-world projects, they must handle assumptions concerning simulation like being abstract, costly, too complicated, and disconnected from operational needs. And there is some truth to these assessments. You need to build models, collect data, conduct the simulation, and afterwards you need to put the results into operation by implementing software and sometimes hardware. All this costs time and money without guaranteed success. If the motivation fails, also the chance to develop new methods that are validated in industry projects gets lost - a problem the authors seem to share with other researchers. The idea of this paper is to have a systematic framework that maps the right Artificial Intelligence (AI) methods to each phase of the simulation life-cycle to increase the usefulness, accessibility, and comprehensibility of simulation for industry. The main outcome is a list of applicable method clusters for each phase to be implemented in a process simulation modeling and execution framework for future applications.» weiterlesen» einklappen

Schlüsselwörter

• Process Simulation
• Literature Review
• AI in Simulation
• Research Agenda

Autoren

Klassifikation

DFG Fachgebiet:
4.43 - Informatik

DDC Sachgruppe:
Informatik