A Workflow Hybrid as a Multi-model, Multi-paradigm Simulation Framework

We propose that workflow software can be coupled with existing simulation frameworks (particularly agent-based ones) to provide three broad benefits: an improved modelling process due to the separation of concerns and rich scheduling syntax; interchangeable human and AI agents at minimal development cost; a common conceptual and software base for multi-model, comparative studies of the same system (including shared, distributed data visualisation). We explain these benefits, providing a proof-of-concept framework implementation and examples from the domain of electricity generation expansion planning.