When disaster strikes, the search for missing flood victims is a daunting task. As robotics researchers who have studied the use of drones in disasters, we know firsthand the challenges of using technology to locate those in need. While drones and AI have shown promise in rapidly determining building and road damage after hurricanes or wildfires, their effectiveness in searching for flood victims is limited.
In this article, we’ll explore the limitations of current implementations of AI in searching for flood victims and discuss the potential for AI-human collaboration to revolutionize this process.
AI’s Potential
Searching for flood victims is a type of wilderness search and rescue that presents unique challenges. Unlike traditional search and rescue operations, where the search area is often confined to a specific terrain or environment, flood victims can be scattered across a wide area, making it difficult to locate them using traditional methods.
However, recent successes in applying computer vision and machine learning to drone imagery for rapidly determining building and road damage after hurricanes or wildfires suggest that artificial intelligence could be valuable in searching for missing persons after a flood.
Developing Computer Vision and Machine Learning Systems
Developing computer vision and machine learning systems for finding flood victims is difficult for three reasons. Firstly, existing computer vision systems are capable of identifying people visible in aerial imagery, but the visual indicators of a flood victim are often very different compared with those for a lost hiker or fugitive. Flood victims are often obscured, camouflaged, entangled in debris or submerged in water, making it challenging for AI systems to accurately detect them.
Secondly, modern AI systems have a high error rate, which can lead to false positives or false negatives. If the system is programmed to overestimate the number of candidate locations in hopes of not missing any victims, it will likely produce too many false candidates, overwhelming search teams and potentially delaying the rescue operation.
Lastly, the lack of high-quality training data for AI systems to learn from is a significant challenge. The rarity of flood events and the complexity of the environment make it difficult to collect and label large datasets for training AI models.
Where AI Falls Short
While AI has the potential to play a significant role in searching for flood victims, it falls short in several areas. One major limitation is the high error rate of modern AI systems, which can lead to false positives or false negatives. Additionally, AI systems lack the contextual understanding and situational awareness that human search and rescue teams possess, making it difficult to accurately detect and locate flood victims.
Furthermore, AI systems are not yet able to adapt to changing environmental conditions, such as shifting debris or changing water levels, which can make it difficult to locate flood victims.
AI-Human Collaboration: The Key to Success
While AI has its limitations, it is not a replacement for human search and rescue teams. AI-human collaboration is the key to success in searching for flood victims. By combining the strengths of AI and human search and rescue teams, we can create a more effective and efficient search operation.
AI can be used to quickly scan large areas and identify potential candidate locations for flood victims, while human search and rescue teams can then focus on verifying the accuracy of these locations and rescuing those in need.
In conclusion, while drones and AI have shown promise in searching for flood victims, their effectiveness is limited by the challenges of flood environments and the limitations of current AI systems. AI-human collaboration is the key to success in this area, and by combining the strengths of AI and human search and rescue teams, we can create a more effective and efficient search operation.
Originally published on https://www.newindianexpress.com/lifestyle/tech/2025/Jul/20/why-drones-and-ai-cant-quickly-find-missing-flood-victims-yet
