Titulo Estágio
Enhancing Fiber Optic Sensor Performance Through Machine Learning
Áreas de especialidade
Engenharia de Software
Local do Estágio
Instituto Pedro Nunes, R. Pedro Nunes Bloco C, 3030-199 Coimbra
Enquadramento
THE COMPANY
FiberSight, a CERN startup focused on innovation for sustainability, aims to ensure reliable and cost-effective technological solutions that empower various industries in monitoring critical parameters in real-time. Our product consists of a temperature and humidity sensor designed to address several issues such as water waste in the public distribution network, optimization of water consumption in large-scale agriculture, or even the detection of infiltrations and structural problems in large constructions such as tunnels and dams.
THE PROJECT
Distributed fiber optic sensors are advanced technological systems that utilize optical fibers as sensing elements to enable real-time monitoring and measurement over long distances. These sensors make use of the intrinsic properties of optical fibers, such as their ability to transmit light signals, to detect various physical parameters including temperature, strain, pressure, vibration, and humidity. Humidity monitoring plays a crucial role in various water-related applications, including agriculture and water leakage monitoring systems. The accurate measurement of humidity levels is of utmost importance in these contexts to ensure optimal resource management. The implementation of a distributed fiber sensor offers a promising solution to reduce water waste, especially considering its scarcity and essentiality for human well-being. By enabling precise and localized monitoring of humidity, this technology can provide real-time data that aids in identifying areas of excessive moisture, optimizing irrigation practices in agriculture, and detecting water leakage in various systems.
Objetivo
Our project aims to leverage machine learning (ML) to enhance the performance of our fiber optic sensor, which detects humidity and temperature changes to identify water leaks. A significant challenge lies in optimizing the hardware components of the sensor. The system requires fine-tuning a large combination of parameters, including:
- Temperature and current of the laser
- Wavelength laser scanning range
- Averages from both optical and electrical signal oscilloscopes
- Spatial resolution
- Pulse width
- Delay times
- Wavelength stabilization times
These parameters must be optimized to achieve a good signal-to-noise ratio, spatial resolution, signal stability, measurement range, and response time. Different parameter combinations may be ideal for specific conditions (e.g., stable measurements) but may perform poorly under large variations. If the hardware parameters are not optimized, signal quality degrades, leading to significant errors, measurements out of range, and false positives.
The main goal of this thesis is to develop ML-based software that adjusts sensor parameters in real-time based on the system's output. To validate this approach, we will use datasets acquired under stable conditions with high-quality measurements. The software should automatically adjust the parameters in response to any variations, ensuring optimal performance.
In terms of computing requirements, our current hardware will be able to perform the tests.
Plano de Trabalhos - Semestre 1
- Understand basic concepts of Fiber Optic Sensors;
- Literature review on ML techniques;
- Develop ML algorithms to learn patterns from fiber optic sensor data and suggest optimal sensor parameters;
- Prepare first intermediate report.
Plano de Trabalhos - Semestre 2
- (CONTINUE) Develop ML algorithms to learn patterns from sensor data and suggest optimal sensor parameters;
- Create protocols to evaluate ML model performance and validate effectiveness in real-world scenarios;
- Implement ML-based software to seamlessly integrate with the product, allowing real-time parameter adjustments.
- Finalize the master's thesis report, submission of document and preparation for final thesis defence.
Condições
The workplace will be at the company's facilities.
There is the possibility to award the internship with a scholarship, according to the candidate’s profile.
Observações
During the application phase, doubts related to this proposal, namely about the objectives and conditions, must be clarified with the supervisors, via email or a meeting, to be scheduled after contact by email.
Orientador
Tiago Filipe Pimentel das Neves
tiago.neves@fibersight.pt 📩