This paper introduces WeatherFormer, a transformer encoder-based model designed to learn robust weather features from minimal observations. It addresses the challenge of modeling complex weather dynamics from small datasets, a bottleneck for many prediction tasks in agriculture, epidemiology, and climate science. WeatherFormer was pretrained on a large pretraining dataset comprised of 39 years of satellite measurements across the Americas. With a novel pretraining task and fine-tuning, WeatherFormer achieves state-of-the-art performance in county-level soybean yield prediction and influenza forecasting. Technical innovations include a unique spatiotemporal encoding that captures geographical, annual, and seasonal variations, adapting the transformer architecture to continuous weather data, and a pretraining strategy to learn representations that are robust to missing weather features. This paper for the first time demonstrates the effectiveness of pretraining large transformer encoder models for weather-dependent applications across multiple domains.

This is the first article of a multi-part series on using Python and Machine Learning to build models to predict weather temperatures based off data collected from Weather Underground. The series will be comprised of three different articles describing the major aspects of a Machine Learning project. The data used in this series will be collected from Weather Underground's free tier API web service. I will be using the requests library to interact with the API to pull in weather data since 2015 for the city of Lincoln, Nebraska. Once collected, the data will need to be process and aggregated into a format that is suitable for data analysis, and then cleaned. The second article will focus on analyzing the trends in the data with the goal of selecting appropriate features for building a Linear Regression model using the statsmodels and scikit-learn Python libraries. I will discuss the importance of understanding the assumptions necessary for using a Linear Regression model and demonstrate how to evaluate the features to build a robust model.

Aircraft maker Airbus is turning to smart industrial drones, data analytics and machine learning to make aircraft inspections easier and faster. One day while working on a shiny new Airbus A350 aircraft, Ronie Gnecco figured it was time to build a better relationship between drones and passenger airplanes. His bold idea to use flying robots for aircraft safety inspections worked so well it has -- among other projects -- it inspired aircraft manufacturer Airbus to move deeper into the industrial drone revolution. Within a couple of years, the company's intelligent unmanned aerial vehicles (UAV) systems could be used for safety inspections at airports around the world, making planes safer with more on-time flight departures. To make that happen, Gnecco said it will require pioneering efforts from technology experts, regulators and airport authorities from around the world.