Medical process automation and improved management efficiency: the Award-Winning projects in the first UPC AI Competition

The emergence of artificial intelligence (AI) has led to an acceleration of many technological processes, which has made it possible to advance in the automation of complex processes, the analysis of large volumes of data, and the development of tools that have a direct impact on people’s health and lives. With the aim of promoting the adoption of AI in the University’s fields of knowledge – architecture, engineering, science, and technology – the first UPC AI Competition was launched, which has awarded projects that have used this tool to improve people’s quality of life or to address business challenges. The Fractus-UPC Deep Tech Hub proposed one of these business challenges, related to patent analysis, to assess how AI could contribute to detecting whether an invention is truly new or whether there are previous scientific texts, a tool that would represent an advance within the intellectual property sector.

A new edition of the competition will open shortly and will once again be sponsored by Fractus through the Fractus-UPC Deep Tech Hub.

AI applied to healthcare receives the most awards
A total of around fifteen awards were presented across two categories: best UPC-wide project, in both individual and collaborative formats, mostly recognizing initiatives focused on the application of AI to healthcare; and best project from each academic school, also in both formats.

The Award for Best Individual Project across the UPC, worth €1,500, went to Rocío del Carmen Ávalos, from the Barcelona East School of Engineering (EEBE), for a project addressing the automation of intracranial electrode contact localization, a fundamental process in the surgical planning of patients with epilepsy. By developing an AI module integrated into the 3D Slicer medical image processing and analysis platform, the time required for this procedure was reduced from four hours per patient to 30 minutes.

In the same category, Miquel López, a student at the Barcelona School of Informatics (FIB), received a runner-up award worth €500. López developed a foundation model based on vision transformers for brain magnetic resonance imaging, trained using the large-scale BrainATLAS dataset and the FreeSurfer tool. The model demonstrated strong performance both in baseline methods and in few-shot scenarios and complex tasks, making it a robust and versatile solution for brain MRI with significant potential for clinical applications.

The Award for Best Collaborative Project across the UPC, worth €3,000, went to FIB students Martina Massana and Maria Sans for developing an AI system designed to predict the presence of bacteria in the bloodstream of patients with hematological cancer and neutropenia, a deficiency of a type of white blood cell that is essential for fighting bacterial infections. This prediction enables earlier clinical decision-making and treatment before infections become critical.

In the construction sector, the student team formed by Aleix Cases and Martí Warda, from the Barcelona School of Civil Engineering (ETSECCPB), received the runner-up award in this category, worth €1,500, for Brinkr, AI-powered software that automates and centralizes companies’ administrative tasks. The tool connects to email and to a folder where documents are stored, while the AI manages orders and invoices and sends alerts when documents are missing or budgets do not match.

The four projects also received awards in their respective categories for the best project from their academic schools.

Academic school awards
In addition to the UPC-wide winners, several of the University’s academic schools also presented awards for individual and collaborative projects.

The winners from each school in the individual category, each receiving €500, were:

• Mauro Filomeno, from the Interdisciplinary Higher Education Centre (CFIS), for the project Online Preference Optimization for Diffusion Models with Classifier-Free Exploration
• Jan Sherzad, from the Barcelona School of Agri-Food and Biosystems Engineering (EEABB), for the project ALGE.AI
• Rocío del Carmen Ávalos, from EEBE
• Marc Rubí, from the Castelldefels School of Telecommunications and Aerospace Engineering (EETAC), for the project Mecalux Pallet Detection
• Víctor Pacheco, from the Barcelona School of Building Construction (EPSEB), for the project Evaluación de métodos de segmentación automatizada en ortofotografías del puerto de Barcelona
• Tomás Sesma, from the Terrassa School of Industrial, Aerospace and Audiovisual Engineering (ESEIAAT), for the project Arquitectura y Desarrollo de Asistentes Conversacionales Inteligentes en entornos específicos. Implementación del backend para el caso de uso: Aeropuerto de Atlanta
• Beiging Zhu Lin, from the Barcelona School of Industrial Engineering (ETSEIB), for the project Efficiency Assessment of Hydropower Units by means of Artificial Intelligence
• Gerard Comas, from the Barcelona School of Telecommunications Engineering (ETSETB), for the project Unsupervised anomaly detection in medical imaging: Brain tumour segmentation in MRI
• Miquel López, from FIB
• Josep Pujolà, from the Barcelona School of Nautical Studies (FNB), for the project Smart-Port-AI

In the collaborative project category, the awards from each academic school, also worth €500 each, went to:

• Damià Bosch, Joel Montemayor and Daniel Vivas, from EEABB, for the project REG.AI
• S.M. Rakib Hasan and Md Sakib UI Rahman Sourove, from EETAC, for the project  TourismBalance: AI-Driven Sustainable Tourism Management
• Oriol Escudero and Biel Puig, from ESEIAAT, for the project SmartLogix: Plataforma de Manteniment Predictiu per a Transelevadors i Transportadors – Intel·ligència Artificial al Servei de la Intralogística
• Aleix Cases and Martí Warda, from ETSECCPB
• Martina Massana and Maria Sans, from FIB