Algorithms and methods for Fiber Bragg Gratings sensor networks

The aim of this work is to design methods and algorithms for a new fully comprehensive cloud platform, which allows to retrieve and analyse data from a set of Fibre Bragg Grating (FBG) sensors. These sensors are able to measure the displacement of a physical object caused by strain or temperature gradients. FBG technology has been widely used in many different fields, like structural health monitoring, preventive maintenance for transportation systems, monitoring of temperature and strain in rotating parts of large machines in power generation plants and more. Today FBG is a cutting-edge technology that allows to precisely measure several physical quantities, with a high reliability in environments subjected to EMI and strong electrical potentials, where the application of standard sensors may be worthless or even dangerous. These sensors require a specific hardware gear, defined as interrogator, which uses a laser source to generate a signal that excites the FBG sensor. A receiver measures the reflected spectrum, which allows to infer the object displacement. There are several manufacturers of interrogator devices, each of them providing a proprietary and non-standard solution to access data. Nowadays the raise of Internet Of Things (IoT) platforms enables a better integration of sensors and data storage platforms. The advantages of an extensive connected sensor network are clear and a dedicated field of research is born in the past decade. However, the lack of a unique method to interface with hardware sensors makes the integration more difficult. The main goal of the research is to create a general purpose open source architecture suitable for any FBG-based system. The software stack provides a lower level interface to a generic interrogator, a cloud data storage infrastructure that remotely stores collected data and a visualisation and analysis framework. An integrated AR/VR system helps interpreting data collected through FBG sensors. Another key point of this work is to set up a general purpose cloud database which allows to integrate metadata from FBG sensors (like their position on the monitored physical system, their current status, etc.). The database must support high data throughput and low latency to comply with real-time data analysis and it is reusable for any FBG-based application. The research has been conducted in collaboration with the DIMEAS and DISAT departments of Politecnico di Torino, as part of the Inter-Dipartimental Center for Photonic technologies (PhotoNext). The platform has been designed to work in different scenarios. Three physical systems will be considered: the first one is the wing of the ICARUS Unmanned Aerial Vehicle, developed by DIMEAS. The main goal is to monitor the bending of the wing during flight. The second scenario aims at monitoring the health status of a bridge model, while the third test is performed onto an oil pipe, in order to monitor the temperature gradients of the structure. All these scenarios require different approaches and have different constraints (for instance the minimum latency between data detection and visualisation, the ability to perform on-line data analysis, etc.), thus demonstrating the flexibility of the proposed architecture.