Sensors, AI and IoT can predict natural disasters
Natural disasters cause catastrophic damage to humans and infrastructure every year, therefore finding a way to reduce this impact has been a key focus.
Detecting natural disasters with IoT, sea-floor sensors, artificial intelligence and machine learning looks to be a promising way forward.
Connected distributed infrastructure is the key to achieving data transmission, which is the foundation of IoT.
Sensor and satellite technology is used to monitor and track data about natural weather events and offer real-time information during and after the occurrence of natural disasters. For example, sensors can monitor roads with damaged power lines and track how quickly a forest fire is spreading.
Once a sensor or satellite detects a potential threat, IoT can be utilised to inform the public about the situation. Local authorities can send alerts through phone apps, voice-controlled devices and vehicle-to-vehicle infrastructures.
Despite advancements in sensor technologies over the years, oceans and seas remain largely unmonitored due to the cost of the infrastructure. However, a new technique could tap into existing networks of subsea cables on the ocean floor to creating an array of environmental sensors.
Telecom cables on the sea floor could be adapted to detect natural disasters including earthquakes and tsunamis, as well as the effects of climate change. Researchers state that this could be made possible by using stretches of fibre optic cables as sensors to enable continuous, real-time monitoring of the conditions.
This has been tested by a team that includes researchers from the University of Edinburgh. They used a 3,600-mile-long subsea cable running between the UK and Canada. The team demonstrated that earthquakes and ocean signals, such as waves and currents, could be detected on individual spans of the cable. 12 sensors were in place along the cable, but future upgrades could allow this to be increased to more than 120.
Using AI and ML
Researchers have begun experimenting with using AI and ML systems to detect natural disasters. AI can analyse huge amounts of information to track the patterns and intensity of weather. If AI is provided with comprehensive, high-quality datasets based on previous events it can use them with real-time seismic data, geographical information and satellite imagery to discover patterns.
Therefore, AI will be able to anticipate natural disasters before they happen. So far, AI can predict earthquakes, volcanic eruptions, hurricanes and tornadoes by researchers providing it with different information. For example, with earthquakes, researchers can give AI systems information from seismic imaging, therefore training them. The AI analyses the data to learn about the patterns of previous earthquakes, allowing it to predict where an earthquake and aftershock might strike.
Researchers are also exploring ways for AI to predict flooding using rainfall records and flood simulations.