Graph Modeling
HFG-TK: Automated Heterofunctional Graph Generation for Interdependent Infrastructure Systems
HFG-TK automates the generation and analysis of heterofunctional graphs so interdependent infrastructure systems can be studied through a common functional representation.
Why HFG-TK
Heterofunctional graph theory is powerful for representing functions, resources, and service pathways across interdependent infrastructure systems, but manually creating those graphs is tedious and difficult to scale. HFG-TK addresses that gap by automating graph construction from infrastructure models, geospatial data, and community asset layers.
Automated Graph Generation
The toolkit parses data from multiple sources, including MATPOWER, OpenDSS and PyDSS, EPANET and WNTR, SWMM and PySWMM, Path4GMNS, and GIS layers for community assets. It converts domain-specific nodes and links into annotated graph objects with layer, function, resource, and spatial metadata.
HFG-TK turns heterogeneous simulator and GIS data into a functional graph representation that can trace how disruptions affect services and communities.
Functional and Geospatial Modeling
HFG-TK represents power transmission, power distribution, drinking water, stormwater, transportation, and community assets in a unified multilayer graph. Hospitals, fire stations, police stations, schools, colleges, grocery stores, residences, and other assets can be connected to required resources and generated services. Spatial references and nearest-neighbor logic help infer interdependencies when explicit dependency data are unavailable.
Analyses Enabled
Once the graph is generated, HFG-TK supports analyses such as disruption propagation, critical asset identification, service accessibility degradation, and time- or distance-based loss of access to resources. These analyses make it possible to connect asset-level failures with human-centered impacts and community resilience questions.
Research Context
HFG-TK is designed for county-scale interdependent infrastructure settings that include electric power, transportation, drinking water, stormwater, and community assets. It supports structural validation, functional feature generation, outage propagation, and degradation in access to services under disruptions. A manuscript describing the toolkit and its disaster-impact analysis workflow is linked as a related output.
My Role
I develop the graph-generation workflow, simulator parsers, geospatial integration routines, analysis abstractions, and resilience metrics that make HFG-TK usable as scientific software rather than a one-off graph construction script.