As part of the Horizon Europe Twinning AIDA4Edge project, the online event “AI Beyond the Cloud – Intelligence at the Edge of the Real World” was successfully organized on Friday, May 8, by our colleagues from the University of Ferrara (UNIFE), one of the advanced partners of the project.

Special thanks go to Riccardo Zese for the excellent organization and coordination of this inspiring event.

The session gathered more than 30 participants and featured a broad spectrum of talks covering Edge AI, intelligent sensing, industrial quality control, traffic analysis, hyperspectral image processing, and sustainable manufacturing. The presentations clearly demonstrated the ongoing transition from centralized cloud infrastructures toward distributed, adaptive, and real-time edge intelligence. During the event, participants had the opportunity to hear several highly interesting presentations:

🔹 A Machine Learning Pipeline to Analyse Multispectral and Hyperspectral Images – Damiano Azzolini presented advanced approaches for ROI segmentation and pixel clustering in spectral image analysis.

🔹 AI-Powered Visual Inspection: Using GANomaly and GradCAM for Industrial Quality – Alice Bizzarri demonstrated the use of GANomaly and Grad-CAM methods to reduce false positives in industrial quality inspection systems.

🔹 Edge AI + Ice Cream Making = Zero Waste Manufacturing – Simon Dahdal showcased an innovative multi-milestone inference strategy for zero-waste manufacturing (ZWM) and adaptive production optimization.

🔹 Dynamic Covariance Estimation in EKF via Deep Learning for Agricultural Vehicle Localization – Andrea D’Antona presented a hybrid approach combining Extended Kalman Filters and neural networks for adaptive vehicle localization under vibrations, uneven terrain, and sensor disturbances.

🔹 Edge Solutions for the Analysis of Traffic Phenomena in Metropolitan Areas –  Francesco Resca presented distributed inference architectures aimed at reducing latency and improving scalability in traffic monitoring systems.

Our impression is that a wide range of highly interesting and diverse research activities is being carried out at UNIFE. Congratulations to all speakers and organizers for creating such an inspiring and multidisciplinary scientific event.

We sincerely thank our colleagues from UNIFE for fostering stimulating discussions on the future of adaptive, efficient, and sustainable Edge AI systems.