Agent-Based Simulation to Measure the Effectiveness of Citizen Sensing Applications—The Case of Missing Children

Citizen sensing applications need to have a number of users defined that ensures their effectiveness. This is not a straightforward task because neither the relationship between the size of the userbase or its effectiveness is easily quantified, nor is it clear which threshold for the number of users would make the application ‘effective’. This paper presents an approach for estimating the number of users needed for location-based crowdsourcing applications to work successfully, depending on the use case, the circumstances, and the criteria of success. It circumvents various issues, ethical or practical, in performing real-world controlled experiments and tackles this challenge by developing an agent-based modelling and simulation framework. This framework is tested on a specific scenario, that of missing children and the search for them. The search is performed with the contribution of citizens being made aware of the disappearance through a mobile application. The result produces an easily reconfigurable testbed for the effectiveness of citizen sensing mobile applications, allowing the study of the marginal utility of new users of the application. The resulting framework aims to be the digital twin of a real urban scenario, and it has been designed to be easily adapted and support decisions on the feasibility, evaluation, and targeting of the deployment of spatial crowdsourcing applications.