Saudi Imam Abdulrahman Bin Faisal University department of computer Science and Saudi Aramco Cybersecurity chair, published in MDPI a study for a solution for Smart Flood Detection to save lives using the integration of AI (Artificial Intelligence), Blockchain and drones.
According to the study, floods pose a serious risk and require immediate management and strategies for optimal response times. The Saudi city of Mecca has been impacted by climate change in the last decade as floods have increased despite the city’s location in the Arabian Gulf, which has a hot and wet climate. According to the General Authority for Statistics in Mecca, since 2010, the average peak rainfall has increased by 350%. Mecca experienced torrential rains on 23 December 2022, at least partly because of its location, surrounded by mountains, causing numerous vehicles to be swept away.
The authors propose a secure method of flood detection in Saudi Arabia using a Flood Detection Secure System (FDSS) based on deep active learning (DeepAL) based classification model in federated learning to minimize communication costs and maximize global learning accuracy.
As per their abstract, “We use blockchain-based federated learning and partially homomorphic encryption (PHE) for privacy protection and stochastic gradient descent (SGD) to share optimal solutions. Utilizing images and IoT data, FDSS can train local models that detect and monitor floods. The proposed FDSS enabled us to estimate the flooded areas and track the rapid changes in dam water levels to gauge the flood threat. This study concludes with a discussion of the proposed method and its challenges in managing floods in remote regions using artificial intelligence and blockchain technology.
The study introduces a drone application that uses blockchain to manage flooding in remote regions safely and in real-time. The framework can be helpful in missions based on both blockchain and IPFS. The proposed architecture of system nodes makes the process more secure by preventing information from being manipulated and enhancing the data analysis capability within the management system. In a blockchain network, the text data is recorded as part of the transaction information that is recorded during transactions. In addition, a visualization platform will allow access to transaction data, making it easier for operators to supervise their operations.
The study offers a scheme that improves the FL system performance by using DeepAL to select the optimal edge nodes and integrating the learned model parameters into a blockchain-based FL scheme to enhance the reliability and security of the FL system. This method is combined with modern cryptography techniques, such as homomorphic encryption, to achieve a high level of privacy and security capabilities.
In natural disasters, UAVs’ real-time data acquisition can prevent harm by controlling operations efficiently. They can be used to obtain aerial photographs and read water levels, wind speeds, and water speeds to predict weather events, prevent disasters, and aid rescues. These complex interactions can be achieved using AI, the computer-based system that executes tasks requiring intelligence.
With AI and machine learning, systems will be able to resist new, sophisticated attacks with shifting characteristics. Drones must be built with a collective machine-learning model integrating all data from IoT devices and webcams that can be sent to the MEC to create an algorithm with strong predictive capability.
The proposed framework assumes that UAVs collect data and MEC servers store it in the blockchain. This includes basic data, such as the device name, MAC address and type, and geographic data, such as latitude and longitude that help MEC servers acquire data. Before data is added to the blockchain, MEC servers verify UAV validity.
The study utilizes the Internet of Drones (IoD) which can help to save many lives during floods and other catastrophic weather events in places that are difficult for people to reach. IoT devices can be used to collect data on the location and status of people in the affected areas, such as their vital signs, to prioritize rescue efforts.
The data will be sent to a central server where deep-learning algorithms will be used to analyze the data and create a rescue plan. The plan will be sent to relevant organizations involved in the rescue efforts, allowing them to provide aid quickly and efficiently to those in need.
In conclusion the study believes that the system has the potential to significantly improve the efficiency and effectiveness of rescue efforts in disaster situations. By utilizing AI, blockchain, and IoT technologies, the system can quickly analyze large amounts of data and provide a comprehensive rescue plan, ultimately saving more lives.
