Title: Towards predictive crowd based transport infrastructure monitoring system

Abstract: To be able to measure relevant data for transport infrastructure monitoring and to obtain maintenance indicators in a crowd sensing-based fashion, a set of requirements (both from hardware and software points of view) needs to be satisfied. Heterogeneity of smartphones and various uncertainties associated with the mainstream \textit{off-the-shelf} hardware combined with the fact that inexperienced participants will take an active part in the data collection process necessitates that the system accounts for dynamicity, uncertainty, unreliability, and heterogeneity of the environment, participants, and technology. Modern Software Development Kits (SDK) and Application Programming Interfaces (API) provided by the mobile Operating Systems allow us to build applications for the smartphones ecosystem and to interact with components, services, and applications. They also provide specifications for hardware components and modules (e.g. sensors, location, and networking modules). Being a system implemented on a heterogeneous and dynamic smartphone platform, our crowd sensing-based transport infrastructure monitoring system satisfies a number of hardware and software requirements. Hence, we propose a detailed and rigorous methodology to research the requirements and describe the overall system architecture including its functionalities and services.