MARCO Project Summary

The project aims to develop a methodology for the analysis of spatio-temporal simulation of the development of the COVID-19 pandemic enabling the construction of a decision support system in  the field of social distancing. The project's implementation requires collecting and processing of  multi-source epidemiological, spatial, demographic, economic, climatic and social data (among  others, official data on the disease and the course of treatment, highly accurate multi-source terrain  models, population distribution, meteorological data, air pollution, residents' mobility, applied  restrictions). The data collected in the repository will enable the knowledge extraction by building  three simulation models using: (i) multi-agent modeling, (ii) deep learning, (iii) Monte Carlo  simulations. The simulation methods used will facilitate the extraction of outbreaks (characteristic  of COVID-19). The results obtained with each of the methods will be compared and calibrated  mutually, based on historical data on the spatio-temporal spread of the COVID-19 pandemic.

Within  the project, the obtained results will enable the development of models and algorithms for the  optimization of recommendations for social distancing, as well as effective geo-visualization methods  for the model simulation of the pandemic's progress. Simulation models developed as part of the  project will also facilitate the testing of diverse scenarios for carrying out policies aimed at  preventing the spread of a pandemic, e.g., the potential outcomes of applying the policies used in  Asia (China, Japan, Taiwan), Europe (Italy, Sweden, Germany), and the US to the Polish conditions.  The results of simulations run using three different approaches will be compared with each other in  terms of economic, social, and health costs, computational efficiency, and prediction accuracy. They  will also be used to build decision models and, as a result, to develop a concept of a decision support  system that utilizes official topographic, demographic, and social data.

The publication of four 140+ research articles will constitute the measurable outcome of the project's implementation. The research will use epidemiological data from the Department of Analysis and  Strategy of the Ministry of Health and data collected in the CENAGIS (Center for Scientific  Geospatial Analyses and Satellite Computations) digital repository of spatial data. Consequently, it  will enable running simulations for the entire country, not just for fragmentary test areas. Thanks to  an interdisciplinary project team comprising employees of the Warsaw University of Technology  (Faculty of Economics and Information Technology, Faculty of Geodesy and Cartography, and the  Centre for Innovation and Technology Transfer Management), the University of Warsaw, the  Medical University of Warsaw, and the Institute of Biotechnology and Antibiotics, it will be possible  to implement a complex project using the synergy effect of diverse medical, geoinformation, IT, and  social knowledge. Access to reliable multi-source data will be possible thanks to the cooperation  with the Ministry of Health, WHO, European Center for Disease Prevention and Control, Ministry  of Health (Italy), Sweden (Public Health Agency of Sweden), Germany (RKI Germany), GB (Public  Health England), and Polish municipal authorities. Scientific collaboration with an expert from  Harvard Medical School/New York University will ensure the comparability of results with the  leading models developed in the USA. Because of the simulation systems developed as part of the  project, it will be possible to model and visualize diverse scenarios for an epidemic spread in the  Polish conditions at a high level of detail and to create generalized simulations for the entire world,  as well as evaluate the results on the basis of the official epidemiological data. The end result of the  project (apart from scientific publications) will be developing a concept of a decision support system  for central and local government administration. Another essential element will also be the proposal  of various data visualization methods that utilize professional cartographic knowledge.