AI-enabled technology to predict rains | Fusion - WeRIndia

Prediction of rain is difficult. It is a complex process. Furthermore, it may not be completely accurate.

The rain may not fall when it is forecasted and it may fall in an unexpected time.

In this context, three engineering students from Chitkara University, Punjab along with their professor designed an Artificial Intelligence enabled device to predict rains in real-time.

If they predict rain falls at 10:00 a.m., then it will fall at that time exactly locally.

Kartik Vij, one of the team members says that they collect information from ground-based nodes which is in real-time. Hence, the prediction is precise and accurate.

Normally information is collected by the Meteorological Department based on satellite image data which is 24 hours old.

The weather is unpredictable in 24 hours due to global warming. In this context, the forecasting of rains by the Meteorological Department may not be accurate.

Another reason for their prediction to become wrong is that they cover the entire city, not just individual villages.

But this team of engineers use local data. They provide various forecasts for every five square kilometres.

The team was inspired to develop AI-enabled technology by the words of Kartik’s mother. One day, she asked him to collect dry clothes from the terrace as it may rain.

Then, he thought that if weather forecasting was in real-time, such uncertainties could be avoided.

Real-time weather monitoring would also help farmers. Then, the team started working on developing technology that could predict rain in real-time.

They noticed signal distortions during bad weather conditions and found that dish TV signals could tell them about the weather.

They installed a device (BhuGoal nodes) between the set-top box and the dish antenna to record the variations in signals.

These variations were fed to AI algorithms to have accurate weather forecasting.

Image Reference: Thebetterindia