Can AI/MI technology really be used? A practical guide to learning through actual industrial application cases
Series Overview
This series is a practical educational content in four chapters that allows you to learn how Materials Informatics (MI) and AI technologies are actually utilized in industry and what results they have achieved, along with specific examples from companies and research institutions.
Not just theory, we provide detailed explanations of executable Python code examples, actual company names and numerical data, and processes to success, allowing you to learn while thinking about "how to apply this to your own research and work."
Features:
- ✅ Actual success stories: 20+ cases including Exscientia, Toyota, BASF, NASA
- ✅ Quantitative data: Specific numbers such as development period reduction from 10→1 year, cost reduction to 1/10
- ✅ Implementable code: 13 fully functional Python code examples
- ✅ Broad fields: Comprehensive coverage from drug discovery, batteries, catalysts, semiconductors to space development
- ✅ Latest trends: Technology trends as of 2025 and vision for 2030
Total Learning Time: 85-100 minutes (including code execution and exercises)
How to Learn
Recommended Learning Order
For those who want to understand the overall picture of industrial applications (recommended):
- Chapter 1 → Chapter 2 → Chapter 3 → Chapter 4 (all chapters recommended)
- Time required: 85-100 minutes
For those interested in specific fields:
- Drug discovery: Chapter 1 only (20-25 min)
- Energy materials: Chapter 2 only (22-25 min)
- Catalysts: Chapter 3 only (20-25 min)
- Other industries: Chapter 4 only (25-30 min)
For those considering implementation in their company:
- Your field's chapter → Chapter 4 (challenges and future directions)
- Time required: 45-55 minutes
Chapter Details
Chapter 1: AI Drug Discovery in Practice - Accelerating New Drug Candidate Discovery 10x
Learning Content
- Challenges in drug discovery processes - 10-15 years, over 200 billion yen cost
- AI drug discovery approaches - Virtual screening, molecular generation, ADMET prediction
- Success stories
- Exscientia × Sumitomo Pharma: World's first AI-designed drug enters clinical trials (12 months vs. traditional 4.5 years)
- Atomwise: Screening millions of compounds in 72 hours for COVID-19, etc.
- Insilico Medicine: Reached Phase 1 in 18 months, cost 1/10
- Takeda Pharmaceutical: Established AI drug discovery unit, image-based drug discovery
- Implementation examples - Drug-likeness checks with RDKit, molecular VAE, binding affinity prediction
Chapter 2: Accelerating Next-Generation Battery Development - From All-Solid-State Batteries to Perovskite Solar Cells
Learning Content
- Challenges in energy materials development - Li-ion limitations, all-solid-state battery barriers
- MI/AI approaches - DFT alternatives, ionic conductivity prediction, Bayesian Optimization
- Success stories
- Toyota: Discovered solid electrolyte candidates in 1/10 of the time
- Panasonic: Predicting 5000 cycles from 100 cycles with LSTM
- MIT: Discovered 12 new Li conductors in 3 months with GNN + Active Learning
- Citrine: Battery optimization for Uber autonomous vehicles
- Kyoto University: Perovskite solar cells exceeding 20% efficiency in 30 experiments
- Implementation examples - Ionic conductivity prediction, MEGNet DFT alternative, battery degradation LSTM, Bayesian Optimization
Chapter 3: Innovation in Catalyst Design - From Reaction Condition Optimization to Novel Catalyst Discovery
Learning Content
- Challenges in catalyst development - Enormous candidate space, multidimensional optimization
- MI/AI approaches - Descriptor design, reaction mechanism prediction, Transfer Learning
- Success stories
- BASF: 5-10% improvement in chemical process yield, achieved in several weeks
- University of Tokyo: CO2 reduction catalyst, discovered novel Cu alloy in 40 experiments
- Shell: Petroleum refining catalyst, 20% improvement in process efficiency
- Kebotix: Fully automated search system operating 24 hours
- AIST: Theoretical prediction of low-temperature ammonia synthesis catalyst
- Implementation examples - Catalyst activity prediction, multi-objective optimization, adsorption energy GNN, Active Learning
Chapter 4: Expansion of MI/AI - From Semiconductors, Structural Materials to Space Development
Learning Content
- Expansion into diverse industrial fields
- Semiconductors/Electronic materials: Intel, Samsung
- Structural materials: JFE Steel, Nippon Steel
- Polymers: Asahi Kasei, Covestro
- Ceramics: AGC, Kyocera
- Composite materials: Toray
- Space/Aerospace: NASA, JAXA
- Closed-loop materials development - Materials Acceleration Platform
- Large-scale data infrastructure - Materials Project, AFLOW, OQMD
- Challenges and future directions - Data scarcity, explainability, gap with experiments
- 2030 Vision - 90% reduction in development period, 5x improvement in success rate
- Implementation examples - Transfer Learning, Multi-fidelity models, Explainable AI (SHAP)
Overall Learning Outcomes
Upon completing this series, you will acquire the following skills and knowledge:
Knowledge Level (Understanding)
- ✅ Can explain 20+ AI/MI application cases in fields such as drug discovery, batteries, and catalysts
- ✅ Quantitatively understand the technical approaches, results, and development periods of each case
- ✅ Grasp the latest trends as of 2025 and the vision through 2030
- ✅ Know the challenges (data scarcity, explainability, etc.) and their solutions for actual implementation
Practical Skills (Doing)
- ✅ Can implement molecular descriptor calculation, VAE, and docking score prediction for drug discovery
- ✅ Can implement ionic conductivity prediction, degradation prediction, and composition optimization for battery materials
- ✅ Can implement catalyst activity prediction, multi-objective optimization, and Active Learning
- ✅ Can implement Transfer Learning, Multi-fidelity, and Explainable AI
Application Ability (Applying)
- ✅ Can select technologies and approaches applicable to your research field
- ✅ Can evaluate industrial implementation cases and judge applicability to your company
- ✅ Can design practical workflows for data collection, model building, and experimental validation
- ✅ Can understand the technology roadmap toward 2030 and plan advance investments
Let's Get Started!
Are you ready? Start with Chapter 1 and explore the world of MI/AI practical applications!
Chapter 1: AI Drug Discovery in Practice →
Update History
- 2025-10-18: v1.0 Initial release (4 chapters, 20+ industrial cases, 13 code examples)
A practical MI/AI utilization guide learned from actual industrial applications!
Disclaimer
- This content is provided solely for educational, research, and informational purposes and does not constitute professional advice (legal, accounting, technical warranty, etc.).
- This content and accompanying code examples are provided "AS IS" without any warranty, express or implied, including but not limited to merchantability, fitness for a particular purpose, non-infringement, accuracy, completeness, operation, or safety.
- The author and Tohoku University assume no responsibility for the content, availability, or safety of external links, third-party data, tools, libraries, etc.
- To the maximum extent permitted by applicable law, the author and Tohoku University shall not be liable for any direct, indirect, incidental, special, consequential, or punitive damages arising from the use, execution, or interpretation of this content.
- The content may be changed, updated, or discontinued without notice.
- The copyright and license of this content are subject to the stated conditions (e.g., CC BY 4.0). Such licenses typically include no-warranty clauses.