Using semantic graph data for high-volume field deployment of IoT devices

Abstract

Internet of things is becoming an increasingly relevant topic and the amount of deployed IoT object is expanding as the industry manages to reduce the production and operation costs. Each of these devices produces data that is stored and harvested at some point, resulting in financial gain or strategic advancements.

In this thesis we focus on making the deployment and monitoring process of new IoT devices a simple task by developing a dedicated deployment tool and seamlessly integrating the devices into a data market where anyone can publish and search for structured data. For that to happen, every device needs to be deployed having associated semantic data and purpose description, which are defined as ontologies. As the IoT devices come in very versatile hardware types and require different steps of configuration, the platform provides extensive support for existing technologies and also for defining custom types of devices and implements a general flow for deploying new hardware.

The discovery of new data and IoT devices is made possible using a central database which stores information in a graph structure and defines the semantics used by every device and all the information required to access them. The graph structure also provides information about how all entities are connected to each other and what capabilities they have.