EPICGen: an experimental platform for indoor congestion generation and forecasting

AbstractEffectively and accurately forecasting the congestion in indoor spaces has become particularly important during the pandemic in order to reduce the risk of exposure to airborne viruses. However, there is a lack of readily available indoor congestion data to train such models. Therefore, in this demo paper we propose EPICGen, an experimental platform for indoor congestion generation to support congestion forecasting in indoor spaces. EPICGen consists of two components: (i) Grid Overlayer, which models the floor plans of buildings; and (ii) Congestion Generator, a realistic indoor congestion generator. We demonstrate EPICGen through an intuitive map-based user interface that enables end-users to customize the parameters of the system and visualize generated datasets.