For the last decade, the most exciting and game-changing technology was the development of artificial intelligence. More information regarding how AI has shaped the world can be found in robots.net. However, AI became unpredictable as developers started applying it to every basic task from the manufacturing process to the comfort in your own home.
Despite AI’s unpredictability, the technology became one of the building blocks of the future, as it’s currently being developed to predict an incoming calamity such as tornadoes, typhoons, and even lightning strikes. Despite the famous saying, “Lightning doesn’t strike the same place twice,” it sometimes does, and it’s one of the causes of a major catastrophe.
Why Do We Need This Kind Of Technology
We need to face the reality that the world we know and love is not doing so well for the last few decades. Environmental problems such as global warming, urban heat islands, ocean acidification, fossil fuels, greenhouse gases, the rise of sea levels, and other problems are what makes it so alarming.
Due to these problems, the world became very unpredictable. Tremors all around the globe, storms that come in unscheduled, and lightning strikes are recently hitting the same place twice. But due to the artificial intelligence’s adaptability to technology, humans now have a fighting chance at survival as they use AI to predict the next catastrophe to avoid fatalities.
Where Artificial Intelligence Comes Into Play
In a university in Switzerland, scientists have come up with an idea to use artificial intelligence in determining where the next catastrophe would hit, especially lightning strikes. The system is relatively cheap and simple as it can predict a lightning strike moments before it hits. However, it only has a range of 30 kilometers in a radius to where the system is.
Machine Learning In The System
Scientists applied techniques of machine learning in the system to have a successful hindcast on where the next lightning is going to hit. The program uses the perception of meteorological parameters for machine learning to take advantage of, to calculate lightning hazards from a distance.
What is Hindcast?
Unlike forecasting, hindcasting is a method for analyzing mathematical models. Moreover, estimated inputs of events that occurred in the past are used in the mathematical model to familiarize the output that matches already known results. If the output of the model is a match to an already known output, then hindcast is correct.
How They Applied Machine Learning For The System
To achieve a successful prediction, a team of scientists programmed their algorithm for machine learning that observes and identifies weather conditions that could end up with a lightning bolt shooting down from the sky. Moreover, the system uses four variables for the algorithm to use, to create a successful forecast:
- Station level air pressure
- Temperature in the air
- Relative humidity
- Average wind speed
After the program has finished with its learning phase, the system would successfully predict new lightning strikes. However, it’s only accurate 80 percent of the time. The researchers behind the project are hoping to use their system for the European Laser Lightning Rod project as an attempt to create a new kind of protection from bolts of lightning.
How This Invention Could Be A Game-Changer
With great technology comes a great price, as materials and human resources became scarce for the last few years. As of now, systems are running slow and very difficult to manage, while it requires external data that is very costly. However, researchers have figured out a way in which it uses data that can be acquired from weather stations.
This all means that we can finally cover distant regions that are out of range of a satellite or radar, and also from places where network communications are inaccessible. Moreover, as data can be accessed easily, predictions are made faster. Therefore, weather alarms can be broadcast prior to a storm being formed.
Training The Machine In Using Available Data
With the development of machine learning, researchers have applied this technology to the system and training it to recognize weather patterns and conditions. For the system’s training, researchers used ten years’ worth of data from various weather stations in both rural and remote areas.
People have said that there’s more chance of winning the lottery than getting hit by a bolt of lightning. However, that doesn’t seem to be the case. There are recent reports that bolts of lightning are hitting the same place twice, and sometimes even crowded places.
However, technology today is only limited to predicting storm clouds as scientists are having a difficult time predicting when and where the next lightning is going to strike. The technology still has its ongoing improvements, but the development of this technology could help save lives and avoid fatal injuries and sometimes environmental catastrophes.