Agent Based Epidemic Modelling (ABEM)

CGA仿真是一个3D建模和仿真工作室，具有将数学建模过程应用于“现实世界”问题并寻找解决方案的经验。我们在模拟的城镇和城市中使用了“基于代理的建模”（ABM）来测试自动驾驶汽车技术的安全性和干预灾难规划。对于这个项目，我们建议使用ABM对一系列情景和个人行为（包括个人和病毒传播）进行建模，目的是预测比目前使用贝叶斯模型（目前为政府的封锁干预提供信息的数学模型）更具体的结果。我们认为，在大流行的这个更成熟的阶段，ABM模型可以提供急需的具体数据，围绕这些数据制定更有针对性的干预措施和程序，以帮助推进封锁议程，并更有凝聚力地为未来的大流行制定计划。这是因为ABM并没有将整个人口建模为一个同质的、蜂群思维的实体，而是假设了特定人群、群体的个体代理，或者在我们最近使用ABM的一个项目中，自动驾驶汽车。在之前的这个项目中，我们为每辆车建模，就像车主去拜访朋友、上班或去看足球一样。换句话说，他们自己做决定，有个人的影响。这是我们对与covid - 19传播有关的行为进行建模的方法，它将使我们能够针对特定活动（如看电影/踢足球）量身定制有针对性的干预措施和方法。我们还可以调查超级传播者在病毒传播中的作用，并考虑知情干预如何帮助护理人员在照顾病人的同时更安全地避免感染。这种方法是创新的，因为它是一种比当前的建模方法（主要倾向于统计）更注重细节的建模方法。与现有的模型相结合，ABEM将有助于为政策制定者提供他们需要的工具之一，以便在疫苗可用之前使社会和经济恢复运转。此次延期将使我们有时间改进我们的病毒模型，将病毒感染率的后编码数据和锁定策略纳入其中，以提高模拟的保真度和真实感。我们将能够从研发和商业的角度探索围绕艺术和体育的特定用例，以及观众的重新接纳。它将使我们能够通过应用于现实世界的问题并与潜在客户一起工作，将ABEM的技术准备水平从5/6提高到7/8。

CGA simulation is a 3D modelling and simulation studio with experience of applying mathematical modelling processes to 'real world' issues and finding solutions. We have used 'Agent Based Modelling' (ABM) in the simulated towns and cities we have created to test the safety of autonomous vehicle technology and for meddling disaster planning. For this project, we propose to use ABM to model a series of scenarios and individual behaviours (both of individuals and viral spread), with the aim of predicting more specific outcomes than is currently possible using Bayesian modelling (the mathematical modelling that is currently informing the Government's lockdown interventions.)We believe that at this more mature stage of the pandemic ABM modelling can provide much needed specific data, around which to create more targeted interventions and procedures to help move the lockdown agenda forward and plan more cohesively for future pandemics. This is because ABM does not model the entire population as one homogenous, hive minded entity but assumes individual agency of specific people, groups, or in the case of a recent project in which we used ABM, autonomous vehicles. In this previous project, we modelled individual cars as if the owners were visiting friends, going to work or going to the football. In other words, making their own decisions, with individual impacts. This is the approach we would take to modelling behaviours relating to Covid19 spread and it would enable us to tailor targeted interventions and approaches to specific activities like going to the cinema/ football. We could also investigate the role of super spreaders on viral spread and consider how informed interventions could help keep carers safer from infection, whilst caring for the sick.This approach is innovative because it is a much more detail orientated approach to modelling than the current modelling approaches (which tend in the main to be statistical). In conjunction with existing modelling ABEM would help give policy makers one of the tools they need to get society and the economy back up and running before a vaccine is available.Effects of Extension for Impact FundingThe extension will allow us time to improve our viral modeling to incorporate post-code data around virus infection rates and the lockdown strategies in-place to improve the fidelity and realism of our simulation.We will be able to explore specific use-cases around arts and sports and the readmission of audiences from both and R&D and commercial standpoint. It will allow us to move the Technology Readiness Level of ABEM from 5/6 to 7/8 through application to real world problems and working alongside potential customers.