Periodic Reporting for period 1 - SoFiN (Software enabled Fiber optic multisensing Network)
Période du rapport: 2022-12-01 au 2024-05-31
The infrastructure for fibre optic sensors has expanded extensively, with a great potential for more efficiently conducting sensing tasks but also carrying data in parallel. However, their use for sensing is very limited due to a lack of supporting technologies. The EU-funded SoFiN project will address current challenges and carry out research to utilise fibre optic sensors for sensing systems, providing great benefits throughout the already installed and extensive fibre infrastructure. To achieve this, the project team will develop an innovative sensor platform along with novel laser types and processing units, all connected to a centralized software monitoring platform, employing intelligent technologies to be newly developed as part of this research project.
The project is based on three key strategic objectives. First is the development of an adaptive, modular and highly integrated photonic multi-sensing system, capable for various sensing applications in a sensor network. This involves the technical design and further development of the flexible sensor platform that can be used for various fiber-based sensing applications. Second the parallel research and development of new types of digital signal processing and cloud connections of the sensor system. In brief this involves the novel signal processing concepts to improve sensing performance and integration of the sensor system into a cloud-based monitoring system. The third objective involves the validation and demonstration of the project developments under the context of end-user needs. For this objective, the system will be used in three typical application use cases to further study behaviour close to real world scenarios.
The uses cases included in the project are: 1) Supervision of powerlines where distributed temperature and acoustic sensing will be used to collect information on the status of powerlines. Also, it is planned to interrogate FBG (Fiber Bragg Grating) arrays to further demonstrate and validate the flexibility of the system. 2) Supervision of telecommunication where information will be collected from the system, on external impacts threatening the fiber infrastructure, e.g. fire, stress, heavy machinery in the perimeter of fiber cables that could risk fiber integrity. Investigations will be also performed to increase security of unmanned network sites. 3) Supervision of water supply network where distributed acoustic sensing will be used to detect leaks via changing sound generated by water flow, but also FBG array-based supervision to detect temperature and strain variation caused by lost water combined with methods of machine learning to detect water loss at an early stage.
The project is based on three key strategic objectives. First is the development of an adaptive, modular and highly integrated photonic multi-sensing system, capable for various sensing applications in a sensor network. This involves the technical design and further development of the flexible sensor platform that can be used for various fiber-based sensing applications. Second the parallel research and development of new types of digital signal processing and cloud connections of the sensor system. In brief this involves the novel signal processing concepts to improve sensing performance and integration of the sensor system into a cloud-based monitoring system. The third objective involves the validation and demonstration of the project developments under the context of end-user needs. For this objective, the system will be used in three typical application use cases to further study behaviour close to real world scenarios.
The uses cases included in the project are: 1) Supervision of powerlines where distributed temperature and acoustic sensing will be used to collect information on the status of powerlines. Also, it is planned to interrogate FBG (Fiber Bragg Grating) arrays to further demonstrate and validate the flexibility of the system. 2) Supervision of telecommunication where information will be collected from the system, on external impacts threatening the fiber infrastructure, e.g. fire, stress, heavy machinery in the perimeter of fiber cables that could risk fiber integrity. Investigations will be also performed to increase security of unmanned network sites. 3) Supervision of water supply network where distributed acoustic sensing will be used to detect leaks via changing sound generated by water flow, but also FBG array-based supervision to detect temperature and strain variation caused by lost water combined with methods of machine learning to detect water loss at an early stage.
During the first 18 months of the project, work already progressed towards designing and developing a flexible sensor platform that can be used for various fiber-based sensing applications. The current research is also focused on the integration of two interrogator principles, named correlation OTDR and OFDR with coherent detection, into one hardware setup and operated purely by software. The detailed laser specifications are already defined as part of the relevant deliverable. The ultra-low phase noise laser sources are being investigated in parallel with the aim to develop an ultra-low phase noise single wavelength laser. The research is currently also looking into a machine learning approach based on reinforcement learning aiming for quantum noise limited phase noise lasers. The use of different types of sensor elements is under investigation. With respect to fiber as sensor element, currently the different fiber types are being considered and the specific requirements but also availabilities of already installed fibers are being determined.
As part of the second key objective the relevant work is already in progress which involves the development of novel signal processing concepts to improve sensing performance and integration of the sensor system into a cloud-based monitoring system. Different methods are currently being investigated and relevant plans are prepared for integrating ML into the SoFiN system. This is also focused on the implementation and optimization of ML algorithms for enhanced fiber sensing. The output of the investigations included in this task will be used for the SoFiN ML and cloud platform. The first version is planned to be delivered by M22 - Sep. 2024. The Digital Twin as a tool representing real sensing elements on a digital basis, is currently under research. Plans are being prepared on the development required to allow a continuous and automated flow of information between the sensor platform and the digital twin representation. The cloud-based monitoring system is under study at the moment to ensure the correct system will be developed to allow cloud connectivity of sensor systems with high importance for efficient processing of sensors data, hosting and executing the machine learning and artificial intelligence algorithms, and seamless integration of the sensors into the overall monitoring system.
For validation and demonstration, extensive lab tests are planned to verify the project goals. This objective was initiated on M13, primarily with the definition of test scenarios. The technical partners and case study leaders already defined the requirements for each particular use case included in the project. The system will be used in three typical application use cases to further study behaviour close to real world scenarios. The performance of the various interrogation formats will be tested with fibers as sensor elements and with arrays of FBGs.
As part of the second key objective the relevant work is already in progress which involves the development of novel signal processing concepts to improve sensing performance and integration of the sensor system into a cloud-based monitoring system. Different methods are currently being investigated and relevant plans are prepared for integrating ML into the SoFiN system. This is also focused on the implementation and optimization of ML algorithms for enhanced fiber sensing. The output of the investigations included in this task will be used for the SoFiN ML and cloud platform. The first version is planned to be delivered by M22 - Sep. 2024. The Digital Twin as a tool representing real sensing elements on a digital basis, is currently under research. Plans are being prepared on the development required to allow a continuous and automated flow of information between the sensor platform and the digital twin representation. The cloud-based monitoring system is under study at the moment to ensure the correct system will be developed to allow cloud connectivity of sensor systems with high importance for efficient processing of sensors data, hosting and executing the machine learning and artificial intelligence algorithms, and seamless integration of the sensors into the overall monitoring system.
For validation and demonstration, extensive lab tests are planned to verify the project goals. This objective was initiated on M13, primarily with the definition of test scenarios. The technical partners and case study leaders already defined the requirements for each particular use case included in the project. The system will be used in three typical application use cases to further study behaviour close to real world scenarios. The performance of the various interrogation formats will be tested with fibers as sensor elements and with arrays of FBGs.
During the first 18 months, the project has overall progressed well and according to expectations in line with DoA. In terms of delivering the needful impact, the project has already accomplished from this stage important results already published or soon to be submitted for peer reviewed scientific publications. Certain project results have been already communicated at invited talks in international scientific events. This in turn shows that all so far progress combined together aims at strengthening even more the impact of this project in several fields, and importantly, the impact in scientific, economic and industrial production or processes. The project developments will be tested and validated at later stages of the project, and a successful overall validation will provide a first solid proof of the SoFiN concept and hence, indirectly also show its promising potential to deliver the expected impact in the above mentioned fields.