Real-time adaptive algorithm, through-wall radar and organ-on-a-chip mark the 16th edition of the Fraunhofer Portugal Challenge
Real-time adaptive algorithm, through-wall radar and organ-on-a-chip mark the 16th edition of the Fraunhofer Portugal Challenge
In November, the auditorium of Fraunhofer Portugal-AICOS, based at UPTEC – Science and Technology Park of the University of Porto, hosted the final of the 16th edition of the Fraunhofer Portugal Challenge, an annual competition for technology-based ideas developed by students from Portuguese universities. The event brought together finalists, researchers, entrepreneurs and representatives from the academic world to celebrate scientific innovation with an impact on society, in an event marked by knowledge sharing and connection to the innovation ecosystem.
The first prize in the Master’s Thesis Award category was awarded to Lara Sá Neves, who developed her thesis at the Faculty of Engineering of the University of Porto (FEUP) and is currently pursuing a PhD in the dual programme at Carnegie Mellon University – Instituto Superior Técnico. Her proposal, iHOMER, presents a pioneering algorithm capable of learning, reorganising information and adjusting classifications in real time, ensuring explainability and accuracy even in highly dynamic scenarios.
In second place, tied, were Gonçalo Martins, with a platform of totally passive sensors that do not require batteries, and Mariana Valente, with the development of a broadband antenna for through-wall radars, capable of detecting people through walls with high resolution, both from the University of Aveiro. The third prize was awarded to Fábio Dias, from the Instituto Superior Técnico, whose work focuses on robust data transmission solutions for organ-on-chip devices, an emerging technology that seeks to improve drug testing and reduce dependence on animal testing.
In the Student Award category, the winner was Gonçalo Lobo, a student at the Faculty of Sciences of the University of Lisbon, who presented an architecture to enable fully encrypted inference in machine learning models using homomorphic encryption — an approach that allows sensitive data to be processed without compromising user privacy.
The afternoon also included a particularly inspiring moment dedicated to entrepreneurship. In a relaxed and open conversation, Pedro Pinto, Head of Business Development – Tech at UPTEC, and José Valente, co-founder and CEO of Azitek, a startup incubated at UPTEC, shared with the audience the paths they have built as founders and leaders of technological projects, reflecting on difficult decisions, turning points and the motivations that drive them. In addition to sharing their testimonials, they were also part of the competition jury.
27 November 2025
