Gaetano Carmelo La Delfa

Marie Skłodowska-Curie  Postdoctoral Fellow at BISITE Research Group

University of Salamanca, Spain

BIO: Dr. Gaetano Carmelo La Delfa is a Marie Skłodowska-Curie Postdoctoral Fellow at the University of Salamanca, Spain, within the BISITE Research Group.

He previously served as a Research Fellow and fixed term Researcher at the University of Catania, Italy, teaching Health Informatics, tutoring several courses, supervising M.Sc. and Ph.D. candidates and contributing to various national and international projects in the context of indoor localization, smart cities, Internet of Things and resilient infrastructures. His current research focuses on optimizing last-mile delivery through machine learning and on developing advanced indoor localization systems.

Dr. La Delfa earned a second-level Master’s degree in Methodologies and Technologies for Developing Applications for Mobile Systems at University of Catania (Italy) in 2012 and a Ph.D. in Computer Engineering at the same university in 2016, with a dissertation on hybrid indoor localization systems. In 2018 he won a scholarship to participate to the advanced 18-month training project “SENTI - Electronic Sensors, Nano Technologies, Information Technology for precision agriculture” at National Research Council - Institute of Cognitive Sciences and Technologies (CNR-ISTC) of Catania, focused on ICT for archiving and processing big data, linked open data, Semantic Web, Machine Learning.

Dr. La Delfa has authored several peer-reviewed publications in top-tier journals such as the IEEE Internet of Things Journal and Computers and Electrical Engineering. He actively contributes to international scientific communities, such as IEEE/ACM NoCArc series, and Blockchain25. He is Guest Editor of the MDPI Electronics special issue “IoT-Enhanced Localization”. Beyond academia, Dr. La Delfa is an active iOS and Android Developer, with various native and cross-platform apps and games on both the Apple App Store and Google Play, downloaded, as of today, millions of times.

Speech Title: Intelligent Logistics for Smart Cities: Optimizing Last Mile Deliveries with AI and IoT

Abstract: With the exponential growth of e-commerce and shifting consumer expectations towards faster and more flexible deliveries, the optimization of last mile delivery logistics has emerged as one of the most complex, expensive and rapidly evolving segments within modern supply chains.  Particularly within urban environments, this critical logistics stage can represent up to 41% of the total logistics costs. The talk explores the main challenges that last mile delivery Introduces and provides an overview of innovative strategies and technological solutions currently being explored to address these challenges. A central focus of the discussion will be the Vehicle Routing Problem (VRP) and its numerous practical variants, which serve as foundational frameworks for optimizing delivery routes. We will briefly survey both traditional optimization methods, such as exact algorithms and heuristic approaches, and explore emerging machine learning-driven techniques. The talk will end by describing smartDelivery, a research project that combines IoT technologies, Machine Learning and Operational Research for more intelligent routing decisions, aiming to build smarter and more efficient urban logistics systems.

Marin Lujak

IEEE Member

King Juan Carlos University, Spain

BIO: Marin Lujak es Investigador Distinguido en la Universidad Rey Juan Carlos, como parte de la iniciativa de excelencia Beatriz Galindo del Ministerio de Universidades. Doctor en Ingeniería de Organización (2010), Universidad de Roma Tor Vergata. Licenciado en Ingeniería Industrial con especialización en automatización y robótica por la Universidad de Zagreb, Croacia en 2006. Obtuvo el título de Master en Ingeniería Empresarial en 2007 y el título de Doctor en Ingeniería de Organización en la temática de la optimización distribuida de los sistemas multiagente  en la Universidad de Roma Tor Vergata, Italia, en 2010. Desde 2011 ha impartido múltiples asignaturas en la Universidad Rey Juan Carlos y en la IMT Nord Europe (IMTNE), Universidad de Lille en Francia tanto de grado como postgrado, principalmente en el ámbito de la Optimización y la Inteligencia Artificial.  EN IMTNE fue coordinador del Programa de Especialización “Logística y Sistemas Inteligentes de Transporte” dentro del dominio Industria y Servicios en el Máster en Ingeniería Informática. Pertenece al Grupo de Inteligencia Artificial y al Centro de Investigación para las Tecnologías Inteligentes de la Información y sus Aplicaciones (CETINIA) de la URJC. Su investigación se relaciona con la Inteligencia Artificial y la Optimización Combinatoria, enfoques distribuidos y descentralizados de coordinación para Sistemas Ciberfísicos. Los escenarios de aplicación incluyen el transporte inteligente y ecológico, gestión de emergencias y coordinación de equipos de robots.

Speech Title: Dynamic and Cooperative Multi-Agent Task Allocation: Enhancing Nash Equilibrium through Learning

Abstract: This paper proposes the Dynamic and Cooperative Multi-Agent Task Allocation (DC-MATA) problem, focusing on individually rational agents in a cooperative organization, which allocate dynamically changing tasks over time. DC-MATA aims at dynamically improving Nash equilibrium over time through learning in this context. Task utilities evolve dynamically, and learning, conducted in rounds, optimizes agents' task selection order to enhance system performance. Our proposed DC-MATA solution approach assigns agents to tasks with highest utility over time and tends towards the Nash equilibrium that aligns with agents' self-interest while improving the gap with system optimum. We propose a priority-sensitive reward function and four action sampling algorithms (ε-greedy, ε-decay, Adapted Simulated Annealing, and Prior Sequence-Aware Sampling - PSAS) leveraging a Markov decision process (MDP) framework. Simulation experiments on our newly proposed GitHub benchmark instances confirm robust performance, facilitating efficient task allocation in the DC-MATA scenario.