Deployment and service of smart sensors in the society: An IoT browsing system with learning capability

Abstract

IoT browser provides a novel way for people to interact with objects through the Internet. Comparing with traditional web browsing environment, IoT browsing environment has some unique features such as the way of interacting with the objects, the importance of spatial-temporal information of objects, and the necessity of resource reuse. In this chapter, we propose a novel IoT browsing system with a learning capability middleware for IoT browser integrated with contextaware services. The system adopts a service-oriented architecture and provides device interoperability, resource reusability and spatial-temporal awareness. More specifically, the system can provide tailored services to meet the preferences of users by using flow-based programming to establish event flows on the IoT browser. Furthermore, learning is introduced to help the users build their event flows by suggesting them the next possible events based on the historical context information of sensors (or smart things). We develop a prototype of the proposed system and demonstrate the applicability of the system in a home browsing scenario. The prototype shows that with the help of the IoT browsing system, heterogeneous devices can cooperate with each other to provide IoT services in accordance with context inference results. In addition, each device can be reused by multiple services with minimum human supervision to reduce hardware and deployment costs.