Parallel Implementation of a Cellular Automata-Based Model for Simulating Assisted Evacuation of Elderly People

Abstract

This paper presents the principles of a Cellular Automata (CA) based model that focuses on assisted evacuation processes of elderly people and its parallel implementation. It is vital in venues that it is ensured punctual and safe evacuation of all people irrespective of their physical or mental condition. Obviously, in cases of emergency, such groups of people need special care in order to abandon an area. They themselves form a special category of moving people but they can be further subcategorized regarding their mobility capabilities, according to their different ailments (mobility disabilities, vision incapabilities, hearing weaknesses or even mental difficulties). The proposed model incorporates such parameters matching elderly people to different mobility groups that are assisted by specialized personnel. Multiple simulation scenarios have been performed that validate the response of the model qualitatively and quantitatively. Moreover, due to the inherent parallelism of CA, the solution of Graphic Processing Unit (GPU) is preferred to further speed up and proceed with the parallel implementation of the presented computational CA model. The aim of such a high performance computing implementation is to transform the proposed CA model to the basic module of an elderly crowd management anticipative system. Consequently, the model leads also to useful conclusions regarding proper venue layouts for efficient evacuation response.