About the Demonstrators
The NAIC project provides access to advanced high-performance computing (HPC) resources and support for researchers, educators, startups, and small to medium-sized enterprises (SMEs) interested in exploring AI technologies. Work Package 7 (WP7) is dedicated to designing and implementing these demonstrators, led by NORCE in collaboration with Sigma2, Simula, Sintef, UiA, and UiB. These demonstrators help illustrate potential uses of AI and offer a good starting point for future AI-driven applications in Norway.
The demonstrators are developed in Python and are designed to be user-friendly, with interactive notebooks that can be accessed and run on Sigma2's infrastructure. This approach ensures that the use cases are easy to understand, and this way even non-AI-experts can develop an intuition for complex applications.
Key Demonstrators in Development
- Climate Indices Teleconnection Analysis Using AI and HPC: This demonstrator explores how AI and HPC can be applied to climate science by analyzing teleconnections—long-distance relationships between different climate indices. Researchers at NORCE and the University of Bergen are developing a model that can predict climate indices by identifying correlations. This approach allows for more comprehensive and automated investigations compared to traditional manual methods. The demonstrator has been successfully developed and is ready for use.
- Learning Dynamical Systems with Hamiltonian Neural Networks: Led by Sintef Digital, this demonstrator focuses on using Hamiltonian neural networks to learn the dynamics of complex physical systems. These networks are designed to respect the underlying physical laws, making them particularly effective for simulating high-dimensional data without requiring extensive computational resources. The model generates synthetic data and allows users to adjust parameters, making it a flexible tool for various research scenarios.
- Improved Wave Predictions Using Machine Learning: This demonstrator aimed to improve wave predictions along the Norwegian coast by using machine learning to model complex wave patterns. Unfortunately, due to data limitations and high model complexity, the project was not completed. However, the insights gained have provided valuable lessons for future projects that might tackle similar challenges.
- Graph Neural Networks for Classifying Fishing Activities: Another ongoing demonstrator involves using Graph Neural Networks (GNNs) to classify fishing activities based on data from ships and catch reports. By analyzing patterns in ship movements and matching them with catch data, the model provides insights into fishing behaviors along the Norwegian coast. The work is in its early stages, with further development required to finalize the model.
- Multimodal Optimization: This demonstrator focuses on optimizing complex functions using genetic algorithms. It is being designed to handle black-box optimization problems that require minimal configuration. The model is still under development, but it has the potential to solve high-dimensional optimization problems across different fields.
Collaboration and Future Plans
The development of these demonstrators is a collaborative effort involving several research institutions, academic- and industry partners. The aim is to create tools that are not just demonstrations but also practical resources that can be adapted and used by different communities, from academia to public administration and industry. By making these tools accessible and interactive, the NAIC project hopes to lower barriers for using AI, encouraging more researchers and organizations to explore how AI can benefit their work. As the project progresses, more demonstrators will be developed, focusing on different areas where AI can offer practical solutions.
Looking Ahead
The Norwegian AI Cloud project is committed to supporting AI research, education, and innovation by providing access to robust computing resources and developing tools that make AI more approachable. By bridging the gap between advanced AI techniques and real-world applications, NAIC aims to foster a vibrant AI ecosystem in Norway that can drive future advancements in multiple fields. Stay tuned for more updates as we continue to develop new demonstrators and expand the possibilities of AI in Norway!