The recent Covid-19 outbreak has had a tremendous impact on the world, and
many countries are struggling to help incoming patients and at the same time,
rapidly implement new public health measures such as lock downs.
Many of these decisions are guided by the outcomes of so-called SEIR
(susceptible-exposed-infectious-recovered) models that operate on a national
level. Within HiDALGO, we are using our agent-based modelling expertise to
develop the Flu And Coronavirus Simulator (FACS), an agent-based model that
approximates the viral spread at the sub national level. FACS incorporates
geospatial data sources from Openstreetmap to extract buildings and
residential areas within the region. The code is developed as a collaborative
effort between HiDALGO and the Brunel Computer Science Department, and is
being updated on a daily basis. The various changes can be viewed in our open
development repository here: https://github.com/djgroen/facs.
Using FACS, we can model Covid-19 spread on the local level, and provide
estimations of the spread of infections and hospital arrivals, given a range
of public health interventions. Such models, once verified and validated, can
help to support local decision-making for an effective response to the
epidemic with the health and social care capabilities.
Within HiDALGO, we have also leveraged our expertise in geospatial processing
to develop tools for extracting buildings from local geographies. These
scripts are general purpose, and can easily be adopted to create building
graphs for further models. You can find the code for this in: