📅 Day 1: Foundations of Autonomous Scientific Discovery
Topics Covered
- Introduction to autonomous scientific discovery
- From traditional research workflows to AI-powered research workflows
- Role of foundation models in scientific reasoning
- AI co-scientists and research assistants
- Literature mining, research gap identification, and hypothesis generation
- Applications in biotechnology, drug discovery, nanotechnology, materials science, and sustainability
Hands-on Activity
Google Colab Activity: Build a simple AI-powered literature insight workflow using Python to analyze research abstracts and extract keywords, research themes, and possible hypothesis ideas.
📅 Day 2: Multi-Agent AI Scientists for Research Automation
Topics Covered
- What are multi-agent AI systems?
- Role-based AI agents: literature agent, hypothesis agent, experiment planner, data analyst, and reviewer agent
- Agentic workflows for scientific problem-solving
- Prompt engineering for research agents
- Tool-using AI agents for databases, papers, code, and scientific analysis
- Reliability, validation, and human-in-the-loop research supervision
Hands-on Activity
Google Colab Activity: Create a basic multi-agent research workflow where different AI agents perform literature summarization, hypothesis generation, and experiment planning for a selected research topic.
📅 Day 3: Self-Driving Research Labs and Future of AI-Driven Science
Topics Covered
- Introduction to self-driving laboratories
- AI-driven experimental design and optimization
- Closed-loop research: plan, test, analyze, improve
- Use of machine learning for experiment recommendation
- Applications in chemistry, materials discovery, pharma, biotech, and energy research
- Ethical, reproducibility, safety, and governance challenges
- Future career and research opportunities in autonomous science
Hands-on Activity
Google Colab Activity: Simulate a self-driving lab workflow using a sample dataset where an AI model recommends the next best experiment based on previous experimental results.
