Summer School 2025
16 - 20 June 2025
Tecnopolo - Building 15, Piazzale Europa, 1, 42124 Reggio Emilia RE, Room M.0.1
Welcome to the Industrial Innovation Engineering PhD Summer School in Applied Artificial Intelligence!
This program is designed for PhD students with limited prior knowledge of Artificial Intelligence (AI) and Machine Learning (ML) who are eager to explore AI's practical applications in their respective fields.
Through a combination of hands-on workshops, expert-led lectures, and collaborative projects, participants will gain a solid foundation in AI techniques and learn how to effectively integrate AI into real-world research and industry challenges. The focus will be on application-driven use cases, ensuring that students will understand the impact of AI in multiple reseach fields.
Program (tentative)
Program Details (tentative)
The role of data preprocessing and AI-driven models for accurate classification (Prof. M. Antonakakis)
In this talk, we will explore the critical role of data preprocessing in enhancing the performance of AI-driven classification models. Effective preprocessing—such as data cleaning, normalization, feature selection, and augmentation—can significantly impact model accuracy and robustness. We will discuss how different preprocessing techniques address issues like noise, missing values, and data imbalance, ensuring high-quality input for machine learning models. Additionally, we will examine how AI-driven models leverage these refined datasets to improve classification performance across various domains. By bridging data preprocessing with advanced AI techniques, this talk highlights best practices and strategies to build more reliable and interpretable classification models.
Introduction to Continual Learning in Neural Network (Prof. S. Calderara)
In the field of artificial intelligence, Continual Learning (CL) has emerged as a critical paradigm aimed at enabling machines to adapt and learn from a continuous stream of data, mimicking the way humans acquire knowledge throughout their lives. CL holds great promise for real-world applications, from autonomous systems that can adjust to changing environments to personalized recommendations that evolve based on user preferences. During the talk, we will introduce the fundamental principles of Continual Learning and key solutions, discussing the concept of catastrophic forgetting and the main rehearsal techniques used to mitigate it in deep learning architectures.
AI in Manufacturing: challenges and use cases (Prof E. Carpanzano)
This talk explores the transformative impact of Artificial Intelligence (AI) in manufacturing, focusing on how AI-driven solutions optimize production, improve quality control, and enhance decision-making. From predictive maintenance that reduces downtime to computer vision for defect detection and smart automation for adaptive production lines, AI is reshaping modern manufacturing. We will discuss key AI techniques, real-world applications, and the challenges of integrating AI into industrial settings. By leveraging AI, manufacturers can increase efficiency, reduce costs, and drive innovation in a rapidly evolving industry.
Advanced manufacturing technologies powered by machine learning and high-fidelity modelling (Prof. P Franciosa)
This talk explores the role of Artificial Intelligence (AI) in production, highlighting how AI-driven solutions enhance efficiency, quality, and decision-making across various industries. From predictive analytics for demand forecasting to automated process optimization and real-time monitoring, AI is revolutionizing production workflows. We will discuss key AI techniques, practical applications, and the challenges of integrating AI into production environments. By leveraging AI, businesses can streamline operations, reduce waste, and improve overall productivity in an increasingly competitive landscape.
How AI has changed robotics (Prof. V. Villani)
AI is revolutionizing the field of robotics by enabling machines to perform complex tasks with increasing autonomy and efficiency. Robotics, traditionally focused on mechanical systems, has expanded to incorporate AI-driven capabilities such as perception, decision-making, and adaptability. Key applications include robotic manipulation, autonomous navigation, human-robot interaction, and collaborative systems. AI techniques, such as machine learning, reinforcement learning, and computer vision, empower robots to operate in dynamic and uncertain environments, handle diverse objects, and learn from interactions. However, challenges remain in areas like real-time decision-making, safety, scalability, and ethical considerations. This convergence of AI and robotics promises transformative advancements in industries ranging from manufacturing and healthcare to logistics and exploration, driving innovation and reshaping how machines interact with the world and humans.
Embedded Machine Learning: Challenges and Opportunities in Deploying AI at the Edge (Dr. B. Dianat)
Embedded Machine Learning is revolutionizing edge computing by enabling devices to process data and make intelligent decisions locally, without relying on cloud connectivity. This talk will explore the integration of machine learning algorithms into resource-constrained embedded systems such as microcontrollers and IoT devices. We will discuss the challenges associated with deploying ML models on hardware with limited memory, processing power, and energy resources. The session will cover optimization techniques like model compression, quantization, pruning, and the use of efficient architectures. Real-world applications and case studies will illustrate how embedded machine learning is driving innovation across various industries. Attendees will gain insights into current trends, tools, and future directions in this rapidly evolving field.
Quantum Computing for Optimization (Prof. G. Bettonte)
This session explores the fundamentals of Quantum Computing for Optimization, combining theoretical insights with hands-on practice. In the theory session (2 hours), we will introduce key concepts of quantum computing, its advantages for solving optimization problems, and common quantum algorithms. In the lab session (2 hours), participants will gain practical experience by implementing and testing quantum optimization algorithms using quantum simulators. This session is designed for those interested in understanding how quantum computing can revolutionize complex optimization tasks across various domains.
Exploring the use of Decision Trees for Inventory Management (Prof. F. Lolli)
Industry 4.0 has revolutionized our lives, introducing transformative technologies such as Additive Manufacturing (AM) and Cloud Computing. In this seminar, we will explore the application of machine learning to inventory management, with a particular focus on leveraging Decision Trees for strategic decision-making. Our discussion will highlight the role of Decision Trees in guiding inventory decisions, especially those incorporating AM-produced spare parts. In particular, we will examine how Decision Trees can support spare parts management in various scenarios that features AM parts, including: Preventive maintenance, Constrained stock systems, Situations requiring qualification tests, Uncertainty in spare part failure rates
The need for explainable and trustworthy AI systems (Prof. M. Lippi)
As AI becomes increasingly embedded in critical domains like healthcare, finance, and policymaking, ensuring its explainability and trustworthiness is essential. This talk will explore why AI systems must be transparent, fair, and reliable, addressing key challenges such as bias, accountability, and ethical concerns. We will discuss methods for improving AI interpretability, including explainable models and post-hoc analysis techniques, and highlight the role of regulatory frameworks in fostering responsible AI deployment. By the end, participants will gain insights into building AI systems that are not only powerful but also aligned with societal values and trust requirements.
AI Application Green and LCA (Prof. A. Ferrari)
This session explores the role of Artificial Intelligence (AI) in promoting sustainability and optimizing Life Cycle Assessment (LCA). AI-driven approaches are increasingly used to enhance environmental impact assessments, improve resource efficiency, and support decision-making for greener solutions. We will discuss how AI can automate and enhance LCA processes by analyzing large datasets, predicting environmental footprints, and optimizing supply chains for sustainability. Real-world applications will be highlighted, showcasing how AI contributes to carbon footprint reduction, energy efficiency, and sustainable material selection. By integrating AI with LCA methodologies, industries can make data-driven decisions toward a more sustainable future.
AI and Machine Learning Concepts and Appications (Prof. A. Varol)
This course introduces AI and ML concepts for PhD students with little or no prior experience. It covers core concepts such as Supervised, Unsupervised, and Reinforcement Learning, along with key algorithms like classification, regression, and clustering. Students will explore real-world applications in engineering and industry while addressing ethical considerations. Through lectures, discussions, and hands-on exercises, this four-hour course provides the analytical skills needed to apply AI techniques in research.
Next-Gen Cyber-Physical Systems: Leveraging Digital Twin and AI (Prof. M. Picone)
This lecture will delve into the integration of Digital Twins, Cyber-Physical Systems (CPS), and Artificial Intelligence, emphasizing their applications in both smart city and industrial environments. We will discuss the fundamentals of these technologies, their role in real-time data analysis, and the use of AI techniques to enhance system performance. Through case studies, we will illustrate how Digital Twins and CPS can optimize infrastructure management, industrial processes, and environmental monitoring. The session will cover challenges, best practices, and future trends, providing a comprehensive overview of this rapidly evolving field.
Students Presentation and Discussion
Participants will present and discuss possible ideas on how AI and ML are already or can be applied to their research program.