Abdullah Canbaz, University at Albany – A.I. Powered Emergency Response and Recovery

On University at Albany Week:  Can artificial intelligence help us during a crisis?

Abdullah Canbaz, assistant professor in the college of emergency preparedness, homeland security and cybersecurity, delves in to find out.

Abdullah Canbaz is an assistant professor in the College of Emergency Preparedness, Homeland Security and Cybersecurity at the University at Albany. His research explores the intersection of security and privacy in networking, network science, and applied data science. Prior to UAlbany, he taught at the School of Sciences at Indiana University Kokomo. He is the founder of Clever AI Technologies, a LegalTech startup that focuses on engineering human rights practices to tailor and expand tech solutions for human rights practice.

A.I. Powered Emergency Response and Recovery

Artificial intelligence is redefining how we prepare for and respond to emergencies.

At our AI and Complex Systems lab, we’re pushing the boundaries of what AI can do when lives are on the line. Imagine a world where, during a disaster, AI-powered systems rapidly analyze social media posts and emergency messages to detect real-time threats, identify hotspots, and provide immediate guidance to keep people safe. This isn’t science fiction – it’s the core of what we’re working on.

Our research focuses on using cutting-edge AI, like Large Language Models (LLMs) such as GPT-3.5 and Llama2, to decode the chaos that often follows crises. We’re teaching these models to sift through millions of social media posts and 911 calls, picking up on emotional language and urgent needs, helping first responders and public safety officials act faster and smarter.

For example, we’ve tested these AI tools during two high-impact events – the global outcry after Zhina (Mahsa) Amini’s tragic death and the devastating earthquake in Turkey and Syria. In both cases, our models could track emotional spikes and pinpoint areas where tensions were escalating, potentially preventing unrest or guiding resources to where they were most needed.

But our work doesn’t stop at just proving these technologies can work. We’re digging deeper, asking tough questions: Can these AI models navigate cultural nuances? Do they perform equally well across different languages? What happens when we combine different models – does that make them even better at understanding human emotions in a crisis?

Our vision is to create AI that’s not just smart but resilient – capable of making critical decisions in unpredictable environments at speeds far beyond human capacity. We’re not just preparing for the future of emergency management; we’re building it.

Read More:
[CBS 6] – UAlbany lab tackles AI’s “hallucinations”, emotion detection and smart devices
[Meat + Poultry] – New York researchers developing faster, easier Salmonella tests
[UAlbany] – An Inside Look at CEHC’s New Artificial Intelligence Laboratory