José Pedro Pinto, Rafaela Timóteo, Diogo Lavado, and Filipa Alves!
On the 22nd of November, at the Library Auditorium of the NOVA School of Science and Technology, this year's best technology-based ideas were announced, and prizes awarded to the finalists. José Pedro Pinto, Rafaela Timóteo and Diogo Lavado took 1st, 2nd and 3rd place respectively. In the new category - Student Award - Filipa Alves had already been announced as the first winner of the category, but it was at the closing event that she shared a little about her idea and the work she has been doing with her mentor, Beatriz Félix, researcher at Fraunhofer Portugal AICOS (FhP-AICOS).
The prize in this new category stands out precisely because of the support and guidance provided by the FhP-AICOS research team, in addition to the €3,000 money prize.
The jury of the event was made up of the Director of FhP-AICOS, Liliana Ferreira, the President of the Scientific Board of Fraunhofer Portugal, Waldir Júnior, and the Head of Department of Fraunhofer IKS, Narges Ahmidi. The panel of invited experts included Professor João José Pinto Ferreira, Dean Master of Technological Innovation and Entrepreneurship - Faculty of Engineering of the University of Porto; Hugo Gamboa, Founder and president of PLUX, Associate Professor at the Physics Department of Nova School of Science and Technology, and Senior Researcher at Fraunhofer Portugal AICOS; and Hugo Silva, Co-Founder of PLUX, Award-winning inventor, researcher, and entrepreneur in the fields of biomedical devices and data science. The opening session was also attended by the Director of NOVA School of Science and Technology, Professor José Júlio Alferes.
Keep reading to find out more about these winning ideas!
1st Place – José Pedro Pinto (University of Aveiro) | New architecture for rotational self-adaptive electromagnetic energy harvester
In the scope of large-scale energy generation to combat climate change, mechanical complexity, initial investments, maintenance costs and productions costs must be significantly reduced. Intermittency and varying energy sources (e.g., varying sea waves’ characteristics) are considered the most relevant problems. Concerning self-powering battery systems, their limited capacity must be overcome to significantly avoid substitution, which is a mandatory requirement in many applications, such as in intracorporal medical devices (e.g. in smart bioelectronic implants). Besides, the market growth of IoT technologies and BSNs has not been as strong as expected, as the use of batteries is still required. Moreover, most battery systems contain chemicals and metals that are harmful to the environment.
Electromagnetic energy harvesting systems emerge as a very promising alternative to solve these impacting problems both for small-scale and large-scale applications, by ensuring reduced maintenance costs, lower complexity designs, autonomous operations, and self-adaptable capability. These generators can efficiently capture mechanical energy from the environment and have the potential to solve both the intermittency problems of conventional renewable energies and the related limitations of battery systems. This research work is focused on the development of a new electromagnetic rotational generator architecture, optimized to capture low-frequency mechanical energy, such that it can be used to electrically supply both small-scale applications (e.g. using human body motion to power biomedical devices) and large-scale applications (e.g. for energy generation using sea waves). A new rotational generator architecture has been developed to increase its energy density and efficiency by including five innovations.
2nd Place – Rafaela Timóteo (IST University of Lisbon) | Augmented reality environment for oncoplastic breast surgery
Rafaela Timóteo developed a novel Augmented Reality (AR) application, BREAST Plus, to assist and improve DIEAP flap surgery pre-operative planning, a critical step of the procedure. BREAST Plus runs on the Microsoft HoloLens 2 headset, and provides relevant and accurate information of the vascular network of the inferior abdominal wall for marking the location of perforator vessels on top of a patient's tummy skin. This is done by displaying a high-fidelity, 3D digital anatomical model of the patient with the perforator vessels clearly visible, that can be aligned with the patient in the operating room (OR). During the development of the BREAST Plus user interface (UI), two breast reconstruction surgeons provided their input on their needs and requirements to perform skin marking tasks in observation and co-design sessions. The final UI contains a report-like menu, with information obtained from computed tomography angiography (CTA) imaging, and provides buttons to hide or show individual elements of the digital model. The model can be easily manipulated with the help of a fine-tuning bounding box, allowing a manual alignment with the patient’s body.
To collect qualitative data on usability, user satisfaction, and preferences, we performed a small user study with three breast surgeons in two DIEAP flap surgeries.
3rd Place – Diogo Lavado (NOVA School of Science and Technology) | Leveraging Group Equivariance To Build Geometric Interpretability For Trustworthy Machine Learning
The inspection and maintenance of energy transmission networks constitute a critical responsibility for transmission system operators, including entities such as EDP, Portugal's electrical distribution company. Traditional methods for inspecting power grid structures have relied on labor-intensive on-the-ground inspections and expensive low-flying helicopter missions, which are resource-intensive and time-consuming. Recent advancements in LiDAR-based inspections have demonstrated the potential to enhance the speed and precision of power grid inspections. LiDAR sensors capture high-resolution 3D point cloud data, enabling maintenance specialists to analyze and prevent potential hazards, such as fires and damage to the electrical system. However, the vast extent of transmission networks poses a significant challenge for timely and efficient inspection. To address this challenge, we take advantage of an innovative mathematical framework that leverages domain knowledge to define tailored operators, which are then integrated into a novel learning agent called SCENE-Net. SCENE-Net is designed to detect power line supporting towers within 3D point clouds, allowing for rapid and automated inspections. My research introduces SCENE-Net as a powerful and accessible power line tower detection solution in 3D point clouds. By combining domain knowledge with neural network techniques, SCENE-Net paves the way for faster, more efficient, and cost-effective inspections of electrical grids, contributing to the prevention of defects, power outages, and forest fires.
Student Award – Filipa Alves (IADE) | An interactive orrery for cognitive stimulation of adults with intellectual disability
The impact of scientific and technological advances on society depends, in many cases, not only on their intrinsic relevance, but also on citizens’ confidence and ability to understand and apply the main associated concepts. People with Intellectual Disability are citizens with full rights in society. Despite this, the importance of basic scientific and technological literacy is often overlooked in this population.
Filipa Alves designed a simplified mechanical model of the sun-earth-moon system – an orrery – specifically for adults with mild to moderate ID. The product consists of a wooden puzzle composed of a set of gears of different sizes that can be interlocked.
Additional elements include a ruler housing the axes to support the gear mechanism, and a round base representing the four seasons of the year.
Her aim is to investigate whether the use of this puzzle can contribute to cognitive and sensory-motor stimulation and increase the intuitive understanding of basic astronomy and mechanical concepts.