Ana Barros (CESE)
“Ana coordinated the EIT programme at INESC TEC since the very beginning, in 2019, with remarkable success. In December 2021, the business plan for 2022 was announced, in which all projects that INESC TEC coordinates or participates in were approved. This success is a result of Ana’s ability, commitment, and dedication. At the end of December, we received the news of the approval of the European project SoTecIn Factory, whose proposal was coordinated by Ana. For these reasons, we believe that Ana Barros should be nominated for December 2021 edition of ‘The Incredibles’. Sadly, Ana ended her connection to INESC TEC at the end of last year, to embrace another professional challenge”.
– CESE coordinators
How do you characterise your professional path at INESC TEC (main achievements, challenges, etc.)?
Diverse and exciting! Due to these aspects, INESC TEC was the institution where I stayed the longest in my professional career (10 years!) and a place where I’d imagine myself working for the next 20 years. As to the main achievements, which were also challenges, I’d highlight: (1) the projects that I managed to bring to INESC TEC, since they provided me the freedom to create my teams and carry out the research that motivates me; (2) the PhD students I supervised (three finished and two will finish this year); (3) research visits abroad (Carnegie Mellon University, in Pittsburgh, and Silicon Valley), crucial to improve as a scientific researcher and to attract new projects and PhD students; (4) the services provided to companies and business associations that often combine the skills of several INESC TEC centres at the service of the community; (5) the inauguration of the iiLab – Industry and Innovation Laboratory; (6) the launch of the EIT Manufacturing programme and (7) the publications, mainly the book Next Generation Supply Chains: A Roadmap for Research and Innovation and the second edition of INESC TEC Science & Society magazine High Value Added, Resilient and Sustainable Industry.
How important are the activities of the EIT programme for INESC TEC? Was it easy to find a balance between the work carried out at INESC TEC and the role of EIT coordinator?
Very important – for three reasons:
- INESC TEC is a member of the EIT Manufacturing association, and it pays an annual fee, which allows it to apply for EIT Manufacturing specific funding (supported under pillar III of the European Commission’s Horizon Europe). Therefore, it’s vital that INESC TEC explores the opportunity to participate in this association.
- EIT Manufacturing funds activities with a high level of technological maturity, i.e., ideal to support the transfer of technology developed within the scope of our projects, as well as the activities of our spin-offs, helping them find the first clients among the 70 members of the association.
- When technologies aren’t mature year, EIT Manufacturing gives access to many partners, with whom one could collaborate in the elaboration of proposals for pillar II of Horizon Europe, dedicated to research and development.
Yes, it was easy to accommodate the role of EIT Manufacturing coordinator at INESC TEC since it’s consistent with the work I was already developing within the scope of European projects.
Which aspects of your job do you enjoy the most?
The diversity of topics and people I work with, the opportunity to contribute to the development of Portugal’s industrial fabric and the intellectual challenge of bringing together the right people to address an industry-related issue and to reach relevant solutions for companies.
How do you comment on this nomination?
Well, I always wanted to be part of “The Incredibles”, so I’m glad I finally did it! Thank you very much.
Daniel Granhão, Guilherme Carvalho and Tiago Gonçalves (CTM)
“The CTM coordinators would like to nominate the researchers Daniel Granhão, Guilherme Carvalho and Tiago Gonçalves for their exceptional performance within the scope of the international competition promoted by the International Telecommunications Union (ITU). In this competition, called ‘TU AI/ML in 5G Grand Challenge’, the researcher reached the first place among the 32 finalist teams of the 16 challenges of the 2021 edition; the final took place in December 2021. The 2021 edition of this challenge brought together close to 1600 participants from 82 countries. The three members of the CTM, PhD students in areas related to electronics and machine learning, explored their skills in neural networks to design a solution for fast and efficient classification of radio modulation formats, acknowledged by an international panel as the best in the competition. This is also an example of a successful multidisciplinary collaboration between different areas of the CTM, as well as a good multi-institution partnership, since the team also included a member from the Polytechnic Institute of Leiria. Finally, we would like to praise the proactivity and initiative of these researchers, who were able to perfectly articulate the demanding activities of the PhD with the effort and dedication that the competition required. We hope that this successful example can motivate other PhD researchers to participate in this type of initiative”. *
– CTM Coordinators
What challenges did you face during the competition?
In this competition, we were challenged to develop a machine learning algorithm to detect modulations in radio signals, a common issue in terms of 5G communications. More specifically, we aimed to develop the simplest algorithm and apply it to real situations, in which energy and computational resources are often limited. A major challenge we faced was the difficulty associated with the compromise between computational resources and time spent in training models like these.
How did you overcome them?
Our team developed a new neural network that, by itself, is much more efficient in terms of resource use. In addition, we explore compression methods to further reduce the algorithm’s complexity. The development of this network was only made possible thanks to CTM computational resources (GPUs), due to the time limitations imposed by the competition deadline.
What is the differentiating factor of the work you’ve developed?
The main differentiating aspect is the great advantage in terms of efficiency that our solution provides, when compared to previous solutions and other ideas submitted to the competition. Our algorithm turned out to be about 2.6x more efficient than the second-best solution proposed
How do you comment on this nomination?
We’re very grateful for this nomination. It’s good to see others acknowledge our work; it motivates us to keep working with the same dedication. We must thank the guiding team that brought us together, motivating and helping us to participate in this competition; we’d also like to thank INESC TEC for providing all resources needed.
* The researcher Daniel Granhão was selected as the team’s spokesperson.