BoM’s map view of wind speed and direction forecast. Credit: Bureau of Meteorology
This project aims to develop a multi-platform graphical user interface (GUI) for real-time weather forecasting, leveraging pretrained AI models. The system will provide forecasts for key meteorological parameters, including wind speed, wind direction, rainfall, temperature, and solar exposure. The student will gain hands-on experience in the end-to-end pipeline of weather forecasting, from data collection and processing to model integration and visualisation. The pipeline involves:
- Data Collection: Aggregating multi-source data, including publicly available weather forecast data, satellite imagery, radar data, and weather station measurements.
- Data Processing: Cleaning, formatting, and integrating heterogeneous datasets to ensure compatibility with pretrained AI models.
- AI Model Integration: Feeding processed data into a pretrained AI model to generate real-time weather forecasts.
- Visualization: Developing a user-friendly GUI to display forecasts across multiple platforms, including a web interface, Android app, and iPhone app.
The project will focus on creating an intuitive and responsive GUI that visualises real-time weather data in an accessible format for end-users. The student will learn to work with diverse data sources, apply AI techniques, and develop cross-platform applications, contributing to advancements in real-time environmental monitoring.
| Student attributes | |
| Academic background | Computer science, data science, software engineering, electrical or electronic engineering, environmental science, physics, or a related field. |
| Computing skills | · Proficiency in Python for data processing and application development.
· (Optional) Basic knowledge of working with APIs and handling JSON, NetCDF, grib, or CSV data formats. · (Optional) Experience with data visualization tools or libraries. · (Optional) Prior exposure to mobile app and web development or frameworks like Flutter is advantageous but not required. |
| Training requirement | · Training on AWS usage and using Boto3 for data retrieval.
· Introduction to API integration, focusing on authentication and handling real-time data from weather-related web services. · Guidance on pretrained AI model integration, including model selection and input/output formatting. · Tutorials on cross-platform development using Flutter, covering web, Android, and iOS deployment. · Workshops on data pre-processing techniques for meteorological datasets, including handling satellite and radar data. |
| Project timeline | |
| Week 1 | Inductions and project introduction |
| Week 2 | Initial presentation |
| Week 3 | Multi-source data collection |
| Week 4 | Data pre-processing |
| Week 5 | Feeding models and gathering raw results |
| Week 6 | User interface design |
| Week 7 | GUI development (web) |
| Week 8 | GUI development (Android, iOS optional if a MacBook user) |
| Week 9 | Final presentation |
| Week 10 | Final report |