Josip Balen of Univ. of Osijek, Croatia

For decades, fire detection has relied on single-point, reactive sensors. This paradigm is fundamentally limited, often resulting in late detection and a high rate of false alarms. This talk presents a paradigm shift, moving from simple detection to intelligent perception through a multimodal, AI-driven approach. Fusion of diverse data streams, from computer vision and thermal imaging to gas and environmental sensors creates a rich, contextual understanding of a potential fire threat. The core of this approach lies in machine learning models trained to recognize the subtle, multi-faceted signatures of incipient fire, long before traditional alarms would trigger. The goal is to demonstrate the framework for developing systems capable not merely of reacting to smoke, but of proactively identifying high-risk conditions, thereby establishing a new era of predictive safety and automated prevention. 

Biography:

Dr. Josip Balen is an Associate Professor at the Faculty of Electrical Engineering, Computing, and Information Technology in Osijek, Croatia. He specializes in programming, mobile app development, and intelligent transportation systems, with recent research in fire prevention and machine learning. An IEEE member since 2008 and Senior Member since 2020, Josip serves as Vice Chair of the IEEE Computer Society MGA Board and advises the IEEE Student Branch Osijek. He founded IEEEmadC, the largest global student mobile app contest, and co-chairs the IEEE International Conference on Smart Systems and Technologies. Josip has published 40+ papers, delivered 30+ talks, and earned numerous IEEE awards, including the MGA Larry K. Wilson Transnational Award and MGA Innovation Award.

Address:Croatia