Telecommunications and Multimedia, Applied Photonics, High-assurance Software and Advanced Computing Systems – these are the four domains that INESC TEC researchers will explore within the scope of the five projects that were approved through the Call for Exploratory Projects promoted by the Foundation for Science and Technology (FCT).
The funding allocated by FCT to each project was a maximum of €50K, an amount nearly fully allocated to the five projects – with budgets approved at over €49K. With a total budget of almost €250K, INESC TEC researchers will work on innovative ideas in exploratory research projects over the upcoming months.
From fibre-optic cables that can support disaster prevention to wireless autonomous aerial networks that can be used in extreme scenarios (natural disasters), as well as the development of more energy-efficient devices, greater data privacy, and a wider adoption of programming languages, several INESC TEC researchers will work for 18 months on innovative solutions for science and society.
Telecommunications and Photonics – how do they relate and how can they support disaster prevention and drive climate research?
There are fibre-optic cables on the seabed that use light to transmit large volumes of data over long distances and at high speeds. To do this, these cables must resort to sophisticated transmission methods. One of these methods uses polarisation, along with the amplitude and phase of a signal, for data transmission. Polarisation is the orientation of the electric field of light, which can be influenced by external factors like phenomena occurring on the ocean floor. Exploring the polarisation of light is the focus of the GEOSENSE project – involving two research domains at INESC TEC (Telecommunications and Photonics) and the Portuguese Institute for Sea and Atmosphere (IPMA). The main objective of the project is to study and understand how polarisation in coherent optical systems is affected in optimal submarine cables, to explore its potential as a global geophysical sensing tool.
This research may make it possible to monitor, in real time, seismic activities, temperature variations or ocean conditions on a scale never achieved before. These aspects will have a significant societal impact by enhancing disaster prevention and preparedness. Moreover, the project will act as a main driver in climate research and the optimisation of resource management, providing a deeper understanding of the planet’s dynamic processes.
“The project aims to develop a simulation model of a long-distance coherent transmission system to evaluate the sensitivity of optical fibres to external disturbances in underwater cables. In addition, we will carry out laboratory tests with a coherent transmission system of at least 100 km in length, subject to different external stress and deformation effects that mimic geophysical phenomena, allowing comparison with real data provided by current telecommunications systems “, explained Susana Silva, INESC TEC researcher, head of the project.
Understanding how different geophysical phenomena impact the optical signal in underwater cables can overcome the limitations of current technologies. The acquisition of this knowledge will be a powerful new tool for global geophysical science, making a significant contribution to scientific research and environmental monitoring efforts.
Another of the project’s innovations is the application of machine learning techniques to interpret the collected data, which may also represent a new research area in Big Data, applied to geophysics.
“The data collected will allow us to identify patterns, detect anomalies and predict trends. All this will be crucial to the development of visualisation tools that help researchers and decision-makers in the effective interpretation of the geophysical information collected”, explained Henrique Salgado, INESC TEC researcher also involved in the project.
The collaboration with IPMA in this 18-month project (led by INESC TEC) reinforces the practical feasibility of the research, ensuring that the results achieved can be applied in real scenarios, like the monitoring of underwater cables in Portugal.
How do you create a drone network controlled by Artificial Intelligence to operate in natural disaster scenarios?
FALCON is the project led by INESC TEC that will develop a network of drones that act as wireless access points – almost as flying routers -, managed by AI agents. This solution will have a significant impact on society by improving the ability to communicate wirelessly in different environments, including extreme scenarios like emergencies and natural disasters.
How will INESC TEC Communications researchers be able to develop this network? By resorting to the semantic analysis of communications between users on the ground and multimodal data – like text, audio and image. These aspects will allow the drone network to perceive the environment, determine the context in which it operates and define the most appropriate configuration for the users.
“This way, the drones will be positioned and reconfigured autonomously and dynamically, without human intervention, to provide wireless coverage and address the communication needs of all users”, explained Rui Campos, INESC TEC researcher responsible for this project.
The solution that will be developed by INESC TEC researchers will allow the creation of stable and autonomous wireless networks, particularly important in scenarios where permanent infrastructures are damaged, overused or inaccessible, promoting the development of smarter territories and cities, and improving the quality of life and safety of populations.
“By exploring the application of Artificial Intelligence and Large Language Models (LLMs) in aerial networks, the solution developed within the scope of the FALCON project will allow a semantic analysis of both the environment and communications, paving the way for responsive and autonomous networks, with the potential to increase the reliability of critical communications infrastructures”, explained INESC TEC researcher André Coelho, who co-leads this project.
INESC TEC has a 10-year history of research in aerial networks. The FALCON project will lead to relevant new scientific results, based on the research carried out so far, and a solution with significant economic and social impact, concerning autonomous and truly intelligent communications networks.
How to make devices more energy efficient?
For 18 months, INESC TEC researchers will join the AIMaCo-V project, working to improve the performance of a specific type of processor called Coarse Grained Reconfigurable Array (CGRA). CGRA is a promising technology used in devices that must be energy-efficient, such as systems found in cars, drones, smart devices, and more. Although this technology already provides some advantages in terms of energy efficiency, it is not yet widely used because its compilation process – i.e., the method by which the written code is transformed to work on the processors – is very time-consuming.
In this context, the researchers will work to overcome this challenge, applying AI techniques capable of accelerating the compilation process, potentially reducing the time from hours to minutes.
“In addition, we will also integrate CGRA processors with RISC-V processors, which are open-source processors and are widely used in low-cost hardware projects. It will be a way to support the community of developers working with open-source hardware”, explained Nuno Paulino, INESC TEC researcher leading the AIMaCO-V project.
During this period, INESC TEC will be able to find solutions to make this technology more practical and efficient, facilitating the use on devices that require good performance, but also low energy consumption.
How do we ensure data privacy when the need for sharing is increasing?
Another approved project is called TSP2Net (Time Series Privacy-Preserving: New Approaches via Complex Networks); over the course of 18 months, INESC TEC researchers in the field of Advanced Computing Systems will work on a solution focusing on the protection and preservation of temporal data privacy. This solution will bring clear benefits in protecting individual privacy, concerning compliance issues with data protection regulations. In this sense, universities, companies and even government entities will be able to use the new methods to share data securely, avoiding the risk of re-identification.
The main goal? To explore and develop new approaches, based on complex networks, to address the challenges associated with privacy in time series, while ensuring a secure exchange of data – especially in areas like Power and Energy. To develop this solution, researchers will create an extra layer of protection, like a privacy-preserving mechanism, representing and sharing the original data as complex networks (i.e., under a completely different data structure). “In addition, we will develop inverse methods – re-transformation – based on networks, capable of generating a new synthetic time series from the original data networks while preserving the essential properties of the data and ensuring the privacy of confidential and re-identifiable information,” explained researcher Vanessa Freitas Silva, who is responsible for TSP2Net.
How will the researchers develop all these elements? “INESC TEC has a structure that allows the dissemination and availability of knowledge in different scientific domains that are crucial to the successful development of this project, especially in the areas of Computer Science and Engineering, Artificial Intelligence and Power and Energy Systems”, explained the researcher. In addition, the institution benefits from several resources that will further advance this project. One example is the access to real-time series data – specifically focusing on Power and Energy – and the existence of application problems to which this project can contribute, such as the secure sharing of this data.
In addition, the project brings innovations to science, opening new possibilities in data security, with an impact in areas like healthcare, economics and AI. By enabling secure data sharing, the project can accelerate new scientific and technological advances.
How to improve software reliability through automatic program repair?
Through the VeriFixer project, INESC TEC researchers will develop techniques for the automatic repair of programs written in programming languages with support for software verification.
Often overlooked by most users, software is an integral part of every digital device that guides our instant communication, medical diagnostics, or transportation. The same happens with applications dedicated to data exchange. The relevance of these devices places a special focus on fault-free software, which is essential to ensure data confidentiality, the correct identification of diseases, or human safety.
However, writing software without flaws, or identifying them, is not a simple task. In this context, Automatic Program Repair emerges as a promising approach to the automatic correction of flaws in software code.
“In critical systems, we require additional guarantees that the software behaves exactly as expected, something that can be obtained through software verification”, mentioned Alexandra Mendes. Several programming languages with support for verification are available, so-called verification-aware languages, and allow this verification process to take place along with the development of the code. When using these to write code, faults are usually indicated by a verifier who warns the programmer when the specification – the precise description of the properties that the program must ensure – is not met. This indicates that the program does not behave as expected.
“Although this approach assists programmers in detecting flaws in the code, repairing them remains a manual task, as existing work in Automatic Program Repair for verification-aware languages is still scarce”, she added. This project aims precisely to help programmers in this repair task.
VeriFixer’s short-term vision is to build an open-source tool that supports users in repairing flaws in verification-aware programming languages and to provide the community with reference datasets with real-world code examples to support new developments in this area. By reducing flaws, errors, and vulnerabilities, verification-aware programming languages prove particularly valuable in critical sectors like healthcare, aerospace, and the financial sector.
Given the exploratory nature of the project – with researchers in the area of high-assurance software -, it focuses on a specific verification-aware programming language, Dafny, which is used by companies like Amazon Web Services. However, the techniques developed through VeriFixer will be adaptable to other contexts and languages.
Developed in partnership with researchers from INESC ID and Carnegie Mellon University, another objective of the project is to encourage a wider adoption of verification-aware programming languages and to reduce the effort to repair programs written in these languages. Ultimately, this will contribute to the development of safer and more reliable software technologies, directly benefiting society.
The researchers mentioned in this news piece are associated with INESC TEC, UP-FEUP, UP-FCUP and U-MINHO.
News, current topics, curiosities and so much more about INESC TEC and its community!
