Engineers from Rice University have developed an advanced system for tracking flooded roads in real-time, using artificial intelligence (AI) and data fusion to enhance awareness of road conditions during floods. This system, called OpenSafe Fusion, is specifically designed for rapid collection and analysis of data from various sources, including traffic cameras, water level sensors, social media data, and traffic speed, to provide more accurate assessments of road flooding.

AI-based technology and data fusion
This innovative system was created in response to the increasing need for reliable tools that can provide up-to-date information on road conditions during urban floods. Existing tools are often limited in their functionality, as they are not primarily designed for monitoring flooded roads but for general flood-related conditions. OpenSafe Fusion, however, uses advanced machine learning algorithms and data fusion to predict road conditions, including the ability of the system to recognize road flooding based on a combination of different data sources.

Testing of the system involved analyzing historical data from Hurricane Harvey in 2017, when the lack of reliable real-time information severely hampered the work of emergency services and residents in Houston. By using this data, researchers simulated flood conditions on around 62,000 roads in the Houston region, and the system successfully tracked conditions on approximately 37,000 road segments, representing a significant improvement in network coverage.

Responsible use of artificial intelligence
One of the key elements of the OpenSafe Fusion system is the "human factor" in the decision-making process. Although the system relies on automated processes and artificial intelligence, researchers have incorporated mechanisms that allow for human oversight of final decisions. This approach ensures that potential errors resulting from fully automated decisions are avoided, which is particularly important in situations that could endanger human lives.

The study also explores how different communities, given available resources and specific needs, could apply this framework for better assessment of road conditions during floods. With increasingly frequent and severe flood events caused by climate change, such a system could play a crucial role in improving traffic safety and responding to natural disasters.

Future steps and plans
Plans for the future include extensive testing and validation of the system in various conditions and environments. Researchers believe that such a system, which utilizes existing data and infrastructure, will become a critical tool in urban areas around the world, especially in communities that frequently face flooding.

The development of the OpenSafe Fusion system is supported by the National Science Foundation (NSF) and the National Academies of Sciences, Engineering, and Medicine through the Gulf Research Program. Although the system is still in the testing phase, preliminary results show significant potential for improving traffic safety and reducing risks during flood situations.

Source: Rice University