On November 15, the Faculty of Medicine of the University of Beira Interior hosted the IEEE Engineering Day, an annual event organised by the Portuguese Section of the IEEE, which celebrates the achievements of engineering in Portugal. This year the focus under the theme “Shaping the Next Generation of Data Centers and Connectivity” brought to Covilhã a reflection on the vital role of engineering in defining the future of digital infrastructures and connectivity.

The event ended with the long-awaited announcement of the winners of the IEEE Portugal MSc Outstanding Awards, a prize that aims to recognise the most innovative and original master’s theses, which explore topics within the technical and scientific areas represented by IEEE, and which is now on its fifth edition.

Nuno Costa, a recent graduate of the Master in Informatics and Computing Engineering (M.EIC), was one of the winners of the award for his thesis entitled “Leveraging Physics-Informed Neural Architectures as Surrogate Models for Space Weather Forecasting”, which proposes models that represent a significant advance for real-time operational space weather forecasting, providing faster and more reliable forecasts, essential for mitigating the impact of space weather on modern technology.

In the realm of space weather forecasting, accurate modeling of solar wind profiles is crucial due to their significant influence on Earth’s magnetosphere and, consequently, on satellite operations, astronaut safety, and communication systems. This thesis addresses the limitations of current magnetohydrodynamic (MHD) simulation tools, which, while effective, face challenges in computational efficiency and predictive accuracy. To overcome these limitations, this research explores the use of advanced machine learning techniques, specifically Physics-Informed Neural Architectures such as Physics-Informed Neural Networks (PiNNs) and Physics-Informed Neural Operators (PiNOs), for creating surrogate models that balance computational speed and physical accuracy.

This research project, which received a 20/20 classification, was supervised by Professor André Restivo (DEI) and co-supervised by doctoral student Filipa Barros (FCUP), about whom he says: “I’m very grateful to them for supporting me throughout this journey, we have demonstrated the prowess of this approach, with our substitute models already being used in operational environments!”.

Nuno was also awarded the Vestas Award 2024, presented at the New Masters Celebration on November 23.

On November 23^rd, as part of the New Masters 2024 Celebration, Nuno Ricardo Teixeira da Costa, a recent graduate of the Master in Informatics and Computing Engineering (M.EIC), took the stage to receive the *Vestas Award, a cash prize of 2,500 euros, which rewards the best dissertations according to criteria of innovation, applicability and positive cost-benefit ratio for Vestas, a world leader in the renewable energy sector for sustainable energy solutions.

Nuno developed the thesis entitled “Leveraging Physics-Informed Neural Architectures as Surrogate Models for Space Weather Forecasting”, which addresses the limitations of current magnetohydrodynamic (MHD) simulation tools, which, while effective, face challenges in computational efficiency and predictive accuracy. To overcome these limitations, this research explores the use of advanced machine learning techniques, specifically Physics-Informed Neural Architectures such as Physics-Informed Neural Networks (PiNNs) and Physics-Informed Neural Operators (PiNOs), for creating surrogate models of the most advanced and expensive wind simulators, that balance computational speed and physical accuracy.

This research, which received a grade of 20/20, was supervised by Professor André Restivo (DEI) and co-supervised by the doctoral student Filipa Barros (FCUP), about whom he says: “I’m very grateful to them for supporting me throughout this journey; we’ve demonstrated the power of this approach, with our surrogate models already being used in operational environments!

Nuno ends his academic career at FEUP as he began it, with distinction, when at the end of his first year he was awarded the “Incentive Prize of the University of Porto”, given to students who finish their first year with the best marks.

But his career wasn’t all about studying. Nuno was an active member of the NIAEFEUP and JuniFEUP student groups, where he had the opportunity to develop extra-curricular activities that contributed to his development, and was a monitor in four curricular units over two years.

*The prize is awarded annually to a graduate of FEUP’s Masters in Electrical and Computing Engineering (M.EEC), Masters in Informatics and Computing Engineering (M.EIC) and Masters in Mechanical Engineering (M.EM), whose final course average and Master’s thesis classification are not lower than 16 and 18 points respectively for the M.EEC and M.EIC, and 15 and 18 points for the M.EM. + info

As part of the New Masters 2024 Celebration, which took place November 23^rd at the José Marques dos Santos Auditorium of the Faculty of Engineering of the University of Porto (FEUP), the Prof. Doutor Raul Vidal/Deloitte Prize was awarded for the third time.

The jury awarded the prize to Gonçalo José Cerqueira Pascoal, a recent graduate of the Master in Informatics and Computing Engineering, who received a cash prize of 3 000 euros.

Gonçalo developed a Master’s thesis that proposes an innovative method, based on deep reinforcement learning, to compose quantum algorithms more efficiently for specific architectures, helping to mitigate the detrimental effects of noise on the reliability of the computations performed. “Noise-Adaptive Reinforcement Learning Strategies for Qubit Routing”, supervised by Professors João Paulo Fernandes and Rui Maranhão, was considered highly innovative, with an academic contribution and practical interest, leading to a subsequent scientific article.

During his four years at FEUP, he was a member of the Associação Tuna de Engenharia da Universidade do Porto (TEUP), an academic group with over 35 years of history that aims to facilitate the integration of students into the university, preserve student traditions and promote the love for music. During this time, he has actively participated in the organisation of various events of a charitable and cultural nature.

On the day of the award ceremony, Gonçalo on this recognition: “(…) after completing my academic career at the Faculty of Engineering of the University of Porto, today I had the honour of being present at the New Masters 2024 Celebration and being honoured with the Prof. Dr. Raul Vidal Award, created by Deloitte. I would like to thank Prof Dr Raul Moreira Vidal, Deloitte, the members of the jury and all the teachers and colleagues who marked my time at FEUP. In future activities, I will strive to continue to apply the same level of dedication, perseverance and rigor that this award seeks to recognise’.

*This award is intended to recognise each year a recent graduate from one of the following FEUP courses, Master in Informatics and Computing Engineering (M.EIC) and Master in Software Engineering (MESW), who has excelled in curricular activities, for the quality and innovation of the work carried out in the field of Software Engineering, and for student support and social and solidarity activities.

On November 8^th, the José Marques dos Santos Auditorium hosted Company’s Day, an event that takes place annually as part of the FEUP Prime programme, of which Águas do Douro e Paiva (AdDP) is a partner since December 2020, and which aims to bring together the dynamics of business, science and academia, among others, in order to value the talent and knowledge produced.

Immediately after the panel “The Intelligent Future” with speeches by Liliana Ferreira (DEI/Fraunhofer Portugal AICOS) and Luis Paulo Reis (DEI/LIACC), the latter presented the “Pitch Challenge“, a challenge in which four PhD students took the stage and described their research work.

Gonçalo Leão, a PhD student of the Doctoral Program in Informatics Engineering, won the challenge with ‘Robotic Bin Picking of Flexible Entangled Tubes‘, a work that is part of a line of research on the robotic manipulation of flexible objects that tend to become entangled. The specific aim of this project is to teach a robotic arm equipped with a 3D vision sensor to remove an object from a disorderly pile of entangled hooks, tubes or cables, so that no more objects are attached to it (a problem known in the industry as ‘bin picking’). To do this, Gonçalo is investigating how we can use supervised learning and reinforcement learning to determine which object the robot should grab, which part of the object to grab, and what movement the arm should make after the grab (for example, to rotate the arm so that an over-grabbed object falls back into the container).

Commenting on the award, which was presented by the Vice President of AdDP, Gonçalo says that the recognition is very important to him because “it represents the valorisation of work that has been time-consuming and very demanding”. The award shows that there is a demand in society and industry for this type of robotic solutions, which are useful in various production systems, such as the manufacture and installation of cables and wires in vehicles and household appliances, and also in recycling systems, where his work can serve as the basis for a selective waste collection system. The winner is grateful for all the support from his supervising team, composed by Professors Armando Sousa, Germano Veiga and Bruno Siciliano, and also from CRIIS, the INESC-TEC centre where he carries out his research, which has provided him with all the resources and support he needs to carry out his work.

The Center for Responsible AI, led by Unbabel, held the Center for Responsible AI Forum ’24 on November 19 at the iconic Casa da Música in Porto. For the first time, the forum brought together AI startups, research centers, and industry leaders to showcase six innovative AI products and discuss the societal impact of artificial intelligence.

The event also saw the announcement of the winners of the first edition of the SPARK Awards, which recognise the best research in Responsible AI by Masters and PhD students. Among the highlights was Cátia Teixeira, the runner-up out of 16 entries with her presentation “Hubris Benchmarking with AmbiGANs”, developed as part of her Master’s thesis in Data Science and Engineering (MECD), in collaboration with Inês Gomes (student of the Doctoral Program in Informatics Engineering (ProDEI)) and Jan van Rijn (Professor at Leiden University).

One of the major limitations of R&D in ML/AI is proper evaluation of models and algorithms. These models are taking an increasingly important role in our society and economy but evaluation still focuses on predictive accuracy, which is well-known to be insufficient. Cátia proposed AmbiGANs, a methodology that generates datasets of ambiguous instances to estimate whether predictive models are overconfident.

The forum also featured presentations by other students from the Department of Informatics Engineering, who explored new methodologies to improve the evaluation of models and algorithms:

Ricardo Inácio (Master in Informatics and Computing Engineering (M.EIC), in progress) with Meta-learning and Data Augmentation for Stress Testing Forecasting Models, presented at Discovery Science;

Inês Gomes (ProDEI student, in progress) with Finding Patterns in Ambiguity: Interpretable Stress Testing in the Decision Boundary, presented at a CVPR workshop;

Luís Roque (ProDEI student, in progress) with RHiOTS: A Framework for Evaluating Hierarchical Time Series Forecasting Algorithms, presented at KDD.

The forum featured keynote speeches from notable figures such as Virginia Dignum (UN AI Advisory Body), Pedro Bizarro (Feedzai), Francisco Pereira (National Institute of Mental Health, USA), André Martins (Unbabel), and Pedro Saleiro (Opnova).

With panels like “AI Regulation – Time to (AI) Act,” the event emphasized the need for ethical and responsible AI usage, ensuring technological advances align with positive societal impact.

Francisca Rocha Gonçalves, a graduate of the Doctoral Program in Digital Media, has seen her project Underwater Ecotones on the podium of the three projects selected for the ‘Bauhaus of the Seas Sails’ artist residencies, an initiative of CAM (Gulbenkian Modern Art Centre) in collaboration with the municipalities of Lisbon and Oeiras.

The ‘Art & Science – A Call to the Sea’ residency, with the Underwater Ecotones project, will be led by Francisca Rocha Gonçalves (PT), in collaboration with Akira Kira, also alumna from the Doctoral Program in Digital Media (PT), Johannes Goessling (DE) and Pedro Frade (PT), and proposes the study of soundscapes, noises and underwater vibrations in direct relation to living species in the Vasco da Gama Aquarium collection. Underwater Ecotones also aims to rethink the museum space, imagining the possibilities and implications of what a museum-aquarium could become in the future.

The three winning collaborative projects ‘have in common the ability to combine art with fields as diverse as science, gastronomy and design, under the banner of sustainability and the relationship with the local marine environment – the estuary, the river and the ocean’. The projects will be developed until the end of December and will culminate in artistic installations that are expected to be open to the public in February 2025.

Francisca’s research work in acoustic ecology focus on the effects of underwater vibrations and particle movement, exploring how anthropogenic noise affects aquatic life. By using sound art and creative technologies, she seeks to raise social and environmental awareness, translating scientific research into accessible artistic expressions. His work offers new perspectives not only for musical compositions, but also for live performances that promote a deeper connection between human beings and the natural world.

Francisca also recently presented at the Boil Festival (Serralves) the installation ± 5,965 Fathoms, developed in collaboration with the digital artist Maotik, which imagines a speculative environment inspired by the depths of the Mariana Trench (located east of the Philippines, it is the deepest place on the planet at 10994 metres). The deep sea exists in a complex of extremes where sound, light and movement are transformed by pressure and isolation. Through immersive soundscapes and visual elements, this AV work invites audiences to experience the unseen world of the deep ocean, where bioluminescence flickers in total darkness and low-frequency vibrations set the tone.

In these depths, light barely penetrates and sound is transformed into strange, reverberating echoes.

The audience was invited to interpret what life and sound might feel like in these extreme environments under immense pressure, where the line between reality and imagination is blurred.

The artist’s portfolio can be viewed on her website.

Carlos Soares, DEI lecturer and researcher in Artificial Intelligence and Machine Learning, took part in a meeting of the Executive Board of Matosinhos City Council on November 6, presenting a seminar entitled ‘AI is dead. Long live AI!’, which served as the basis for a discussion on the opportunities and risks of this technology in the context of public and local administration, but also in society in general.

“Understanding the potential impact of the changes brought by artificial intelligence on the creation of value in organisations and improving the ‘digital IQ’ of the managers involved is the only way to get to grips with the essence of the new concepts of artificial intelligence before they are integrated into processes, allowing governance and management bodies to be more effectively involved in the issues of digital transformation supported by this technology,” says Carlos Soares.

During the session, several questions were raised that allowed the Matosinhos City Council Executive to present the potential, but also the challenges, of applying AI in organisations:

– The current state of AI and the hype surrounding it: recent advances in AI have brought the field into the mainstream media spotlight, alternately heralding a better life for all and the end of the world. The aim is to clarify the difference between hysteria and reality about these advances;

– How LLMs can add value to organisations: the most visible face of progress in AI are Large Language Models (LLM), popularised in systems such as ChatGPT. The aim is to understand what they really are, what value can be gained from using them, but also what their limitations and risks are;

– How to extract value for organisations with other data: LLMs represent a very limited part of the potential of AI, despite their public profile. The aim is to illustrate the potential application of other AI techniques, often with greater potential value than LLMs.

The DEI lecturer adds that “the aim is to present the concepts in a way that is accessible to people who don’t have a technical profile, but at the same time sufficiently in-depth for them to know the general principles of how systems like ChatGPT work”.