Graph theory is the bedrock for modern research into complex systems. Specifically, routable directed graphs are quintessential tools for modelling and analyzing a wide range of flow problems in transportation, logistics, and energy, to name a few.
Analysis of transport networks takes place on multiple (spatiotemporal) scales. One of the key challenges is maintaining a set of mutually consistent directed graphs that represent the same network with all relevant characteristics and components on different scales.
In this research, we will investigate and develop approaches to derive such multi-scale graph representations automatically from data using a combination of topology- and data-driven methods.
Topology-based methods use attributes and characteristics of the network (connectivity, edge weights and priorities), whereas data-driven methods involve characteristics (e.g. traffic flows, travel times) of the physical process using the network.
It is in the combination of these approaches that scientific contributions are needed and possible. The result will be an open-source suite of tools (prototypes, demos) and two (or more) paper(s) describing the methods and algorithms.