The AI4Nuclear Summer School is designed to support the “new skilling” of individuals in artificial intelligence to enable this technology to be exploited by the nuclear industry. Organised by the AI4Nuclear sub-group of the Nuclear Institute Digital Special Interest Group, it is suitable for:
Led by David Smeatham Independent Engineering and Regulatory Consultant, Ennuvo Limited
Track 1 is aimed at leaders within nuclear organisations looking to apply AI to solve challenges within their organisations. The hackathon will focus on how to enable an innovation led approach to overcome organisational and practical barriers to enabling AI within nuclear organisations. The session is particularly targeted at decision makers and leaders within the nuclear sector however, organisations delivering AI solutions are welcome to attend in a supporting capacity.
Day 1 (AM) (both tracks)
Presentations from leaders and RTO’s on the benefits and challenges of AI.
Day 1 (PM)
- Deployment of innovation – presentation
- What does effective governance of AI look like – presentation
- Group workshop – How does this align with participant’s own organisational arrangements
- Group workshop – Barriers and opportunities and threats & success? Strengths, Weaknesses, Opportunities and Threats?
- At the close of Day 1 the facilitators will help the group identify a number of key themes from the barriers and opportunities exercise for focus on day 2.
Day 1 Evening - reception and networking (both tracks)
Day 2 (AM)
- Workshop addressing key barrier / opportunity #1 through collaborative group work
AI Practitioners Deep Dive “Unlocking AI for Engineering Excellence ”
Led by Prof Nawal Prinja and the Hartree Centre team, the purpose of the track will be to unlock your potential to apply AI within the nuclear sector to de-risk engineering projects.
Day 1 (AM) (both tracks)
Presentations from leaders and RTO’s on the benefits and challenges of AI.
Day 1 (PM)
- Hands-On AI Tools: Overview of the workshop toolchain Anaconda, Python, TensorFlow, Pandas, Jupyter Notebook and scikit-learn. TLearn how to preprocess data, build models, and evaluate performance.
- Practical Applications: Few simple examples to run and learn. Explore real-world examples: Go through some engineering examples.
- Extracting Value from Data: Classification and establishing trends
Day 1 Evening - reception and networking (both tracks)
Day 2 (AM)
- Data preprocessing: Data wrangling to clean, transform, and enrich your datasets. Feature engineering: Uncover hidden patterns.
- Hands-on exercises: Implement AI algorithms. Collaborative projects by working in teams to solve engineering challenges.
- Physics Informed Neural Network (PINN): Show how physics can be added to neural networks to solve engineering problems that comply with the laws of physics.