Forestry 4.0 and Industry 4.0: Use case on wildfire behavior predictions

Forest industries deserve special attention due to relations between environmental impact and social and economic development. The increase of forest fires caused by the untenable exploitation has motivated the application of concepts such as Industry/Forestry 4.0 and Internet of Forest Things (IoFT) towards improving the performance of current supply chains and assuming an environmental responsibility. This research focuses on the application of IoFT for the prediction of wildfires behavior and proposes a semantic platform for heterogeneous IoFT data aggregation that grants interoperability through semantic technologies. The dataset considered climatic-and vegetation-related data gathered by Brazilian government sensors and satellite information on fires, and Machine Learning predicted the areas affected after a fire event. Both platform and predictions were validated and Random Forest predicted the area with 89% accuracy, showing better performance than Deep Neural Network, with 79%.

Automated summary

This research focuses on the application of Internet of Forest Things (IoFT) in predicting wildfire behavior, and proposes a semantic platform for aggregating heterogeneous data and achieving interoperability through semantic technologies. By using machine learning techniques, the study predicted the areas affected after a fire event based on climatic-and vegetation-related data gathered by Brazilian government sensors and satellite information. The validation showed the effectiveness of the proposed platform and predictions.