On the Deployment of IoT Systems: An Industrial Survey

Internet of Things (IoT) systems are complex and multifaceted, and the design of their architectures needs to consider many aspects at a time. Design decisions concern, for instance, the modeling of software components and their interconnections, as well as where to deploy the components within the available hardware infrastructure in the Edge-Cloud continuum. A relevant and challenging task, in this context, is to identify optimal deployment models due to all the different aspects involved, such as extra-functional requirements of the system, heterogeneity of the hardware resources concerning their processing and storage capabilities, and constraints like legal issues and operational cost limits. To gain insights about the deployment decisions concerning IoT systems in practice, and the factors that influence those decisions, we report about an industrial survey we conducted with 66 loT architects from 18 countries across the world. Each participant tilled in a questionnaire that comprises 15 questions. By analyzing the collected data, we have two main findings: (i) architects rely on the Cloud more than the Edge for deploying the software components of loT systems, in the majority of the loT application domains; and (ii) the main factors driving deployment decisions arc four: reliability, performance, security, and cost.