FloodBot in Smart City
Led by Nirmalya Roy, Ph. D.
Nirmalya Roy is currently an Associate Professor and Graduate Program Director in the Department of Information Systems at University of Maryland Baltimore County. He directs the Mobile, Pervasive and Sensor Computing Group (MPSC). He was a Clinical Assistant Professor in the School of Electrical Engineering and Computer Science at Washington State University from January 2012 to June 2013. Prior to that, he worked as a Research Scientist at Institute for Infocomm Research (I2R), Singapore from 2010 to 2011. He was as a postdoctoral fellow in Electrical and Computer Engineering Department at The University of Texas at Austin from 2008 to 2009. He received his Ph.D. and M.S. in Computer Science and Engineering from The University of Texas at Arlington in 2008 and 2004 respectively. He did his Bachelors in Computer Science and Engineering from Jadavpur University, India in 2001.
Project:
Storms and floods cause 70% of the world’s natural disasters. These natural disasters affect on average up to 200 million people in a year, with economic losses between US$50 billion to US$100 billion annually. PI Roy and his team have been investigating how to effectively use multimodal data sources from wireless sensor and social networks such as tweets to improve the resiliency of natural disasters. The project is building predictive data analytics techniques to forecast the severity of flash floods and investigating the appropriateness, and representation of micro-blogging services, and feeds from social network sites associated with real-time flash flood events. The project is augmenting the multi-modal sensor-based data analytics model with human generated facts and observations from micro-blogging services and engineering its real deployment for detecting and assessing the emergent flood situation in a severely flash flood prone area, Ellicott City in Maryland. The team is collaborating with local government officials to deploy and test the proposed flood monitoring and warning systems in Ellicott City. The deployment and adaptation of such early warning systems are on the rise, but their integration with today’s automatic speech recognition system and voice assistance is limited. Motivated by this, we envision that in case of a flash flood event the community would be interested to know about the potential danger in-place of their interest just by asking a simple question. A just-in-time short and relevant answer about the current situation instead of searching a news article or scanning through social media feeds would be more helpful. Therefore, we plan to investigate a real-time flood monitoring chatbot, FloodBot which can assimilate contextual information and infuse the conversational AI power of using various deep learning models. The REU students will be directed to work on this real-time flood monitoring chatbot project and help collect, annotate and visually parse images from potentially hazardous areas.