Using Agent-Based Models in Contagion Modelling

by Healthcare

Traditionally, predictive models cannot be created very quickly impacting how efficiently governments, local authorities and businesses can make data-driven decisions. Coronavirus (COVID-19) is affecting thousands of people and understanding its spread is an extremely complex scientific and engineering problem that must be addressed so that governments can take necessary steps as quickly as possible. We have created an agent-based model of COVID-19 that demonstrates the risk of infection and transmission across the UK. Over the course of three days, we’ve designed the initial agent-based model involving 100,000 agents (people) but could be scaled to mimic the entire population if required. Currently, we don’t know the R0 number for COVID-19, and in the context of complex, large-scale simulations, small errors become dramatically magnified. If one part of the framework is misjudged or data is input incorrectly, the repercussions can be significant – especially if decisions are made based upon the inaccurate output. Indeed, as it stands data sources for such an agent-based model are diverse and create a complex picture, including virus genomes, crowdsourced data and social data. One way around this is to create models which don’t rely on these inputs. The Aether Engine simulation focusses on movement patterns to try and isolate specific scenarios. Focussing on two different hubs – Liverpool and London – the simulation shows how the pathogen might spread from different locations along the UK road network.

Creating An Agent-Based Model Mapping The Potential Spread of COVID-19

This simulation runs in super-real-time and can be quickly re-run as new data becomes available as well as scaled up or down depending on the variances being tested. Important Note: This agent-based model is for illustrative purposes only and does not accurately reflect the prospective spread of the pathogen.
Multi-agent system simulations could drastically increase a model’s flexibility in comparison to traditional techniques. This initial agent-based model could be scaled upwards, both in regards to the number of agents, but also introducing additional layers of complexity, including transport infrastructure data to further sociodemographic information. Contagion modelling has, until now, been a theoretically interesting but practically limited area. Prior restrictions had rendered it difficult for organisations to implement appropriate measures in time. Nonetheless, if we are able to harness the increased data and computational power we can build out accurate, scalable models.  Update: We are delighted to announce that we are renewing our partnership with the Francis Crick Institute. The project will combine analysis of person-to-person interaction with insight into how COVID-19 transmits within an individual, providing a complete picture of the pathogen’s spread. Find out more here.