Adopting and Embedding Machine Learning Algorithms in Microcontroller for Weather Prediction

Abstract

Maritime journeys are significantly depending on weather conditions and so meteorology have ever had a key role in maritime businesses. Nowadays, the new era of innovative machine learning approaches along with the availability of a wide range of sensors and microcontrollers, creates increasing perspectives for providing onboard reliable short-range forecasting of main meteorological variables. The main goal of the current study is to propose a machine learning algorithm, which will be coded into a microcontroller and will be able to predict in short-term the wind speed weather conditions on board of the boat. A regression machine learning algorithm was chosen so that to require the smallest amount of resources (memory, CPU) and to be able to run in a microcontroller. The method was coded suing a powerful programming platform for microcontrollers namely the Zerynth studio. The proposed method was tested on real weather data recorded during a ship journey and its efficiency is proven based on a number of error metrics.