Materials Genomics
Unit 13: Physical Constraints, Trust, and Integration Outlook

Prof. Dr. Philipp Pelz

FAU Erlangen-Nürnberg

01. Title: Physical Constraints, Trust, and Integration Outlook

• Frame the unit in the end-to-end materials discovery workflow and state the decision problems it addresses.
• Applied anchor: constraint penalty ablation.
• Book anchor: [Neuer 6.1–6.3].

02. Learning objectives and expected outputs

• State measurable outcomes (what students can explain, implement, and critique by the end of the unit).
• Applied anchor: model card template.
• Book anchor: [Neuer Ch7].

03. Recap from previous unit and dependency map

• Reconnect prerequisite concepts from earlier units and make dependency assumptions explicit.
• Applied anchor: sensitivity analysis report.
• Book anchor: [McClarren Ch11+12].

04. Why this unit matters for materials discovery

• Motivate with a realistic failure/success scenario from materials discovery practice.
• Applied anchor: failed recommendation post-mortem.
• Book anchor: [Murphy model selection].

05. Reading map and chapter anchors

• Map slide blocks to the key book chapters so students can pre-read and post-review effectively.
• Applied anchor: deployment readiness checklist.
• Book anchor: [Bishop model limits].

06. Why constraints are essential in scientific ML

• Explain why constraints are essential in scientific ml using one concrete materials example and one common failure mode.
• Applied anchor: constraint penalty ablation.
• Book anchor: [Neuer 6.1–6.3].

07. Hard vs soft physical constraints in optimization

• Compare hard vs soft physical constraints in optimization using one concrete materials example and one common failure mode.
• Applied anchor: model card template.
• Book anchor: [Neuer Ch7].

08. Constraint examples: charge neutrality, stability bounds, symmetry

• Diagnose constraint examples: charge neutrality, stability bounds, symmetry using one concrete materials example and one common failure mode.
• Applied anchor: sensitivity analysis report.
• Book anchor: [McClarren Ch11+12].

09. Penalty-based constrained learning

• Apply penalty-based constrained learning using one concrete materials example and one common failure mode.
• Applied anchor: failed recommendation post-mortem.
• Book anchor: [Murphy model selection].

10. Physics-informed loss design concepts

• Define physics-informed loss design concepts using one concrete materials example and one common failure mode.
• Applied anchor: deployment readiness checklist.
• Book anchor: [Bishop model limits].

11. Hybrid model architecture: mechanistic + data-driven

• Explain hybrid model architecture: mechanistic + data-driven using one concrete materials example and one common failure mode.
• Applied anchor: constraint penalty ablation.
• Book anchor: [Neuer 6.1–6.3].

12. Constraint violation diagnostics and monitoring

• Compare constraint violation diagnostics and monitoring using one concrete materials example and one common failure mode.
• Applied anchor: model card template.
• Book anchor: [Neuer Ch7].

13. Explainability levels for different stakeholders

• Diagnose explainability levels for different stakeholders using one concrete materials example and one common failure mode.
• Applied anchor: sensitivity analysis report.
• Book anchor: [McClarren Ch11+12].

14. Sensitivity analysis for trust building

• Apply sensitivity analysis for trust building using one concrete materials example and one common failure mode.
• Applied anchor: failed recommendation post-mortem.
• Book anchor: [Murphy model selection].

15. Causality caution: association vs mechanism

• Define causality caution: association vs mechanism using one concrete materials example and one common failure mode.
• Applied anchor: deployment readiness checklist.
• Book anchor: [Bishop model limits].

16. Model cards and uncertainty statements for deployment

• Explain model cards and uncertainty statements for deployment using one concrete materials example and one common failure mode.
• Applied anchor: constraint penalty ablation.
• Book anchor: [Neuer 6.1–6.3].

17. Reproducibility package: data, code, splits, environment

• Compare reproducibility package: data, code, splits, environment using one concrete materials example and one common failure mode.
• Applied anchor: model card template.
• Book anchor: [Neuer Ch7].

18. Failure taxonomy for materials ML systems

• Diagnose failure taxonomy for materials ml systems using one concrete materials example and one common failure mode.
• Applied anchor: sensitivity analysis report.
• Book anchor: [McClarren Ch11+12].

19. Ethical and safety dimensions in accelerated discovery

• Apply ethical and safety dimensions in accelerated discovery using one concrete materials example and one common failure mode.
• Applied anchor: failed recommendation post-mortem.
• Book anchor: [Murphy model selection].

20. What ML can discover vs what requires experiment

• Define what ml can discover vs what requires experiment using one concrete materials example and one common failure mode.
• Applied anchor: deployment readiness checklist.
• Book anchor: [Bishop model limits].

21. Human-in-the-loop validation workflows

• Explain human-in-the-loop validation workflows using one concrete materials example and one common failure mode.
• Applied anchor: constraint penalty ablation.
• Book anchor: [Neuer 6.1–6.3].

22. Integration with high-throughput and lab pipelines

• Compare integration with high-throughput and lab pipelines using one concrete materials example and one common failure mode.
• Applied anchor: model card template.
• Book anchor: [Neuer Ch7].

23. Decision governance under uncertainty and constraints

• Diagnose decision governance under uncertainty and constraints using one concrete materials example and one common failure mode.
• Applied anchor: sensitivity analysis report.
• Book anchor: [McClarren Ch11+12].

24. When to reject model recommendations

• Apply when to reject model recommendations using one concrete materials example and one common failure mode.
• Applied anchor: failed recommendation post-mortem.
• Book anchor: [Murphy model selection].

25. Robustness checks before deployment

• Define robustness checks before deployment using one concrete materials example and one common failure mode.
• Applied anchor: deployment readiness checklist.
• Book anchor: [Bishop model limits].

26. Case: constrained surrogate for stability-aware screening

• Explain case: constrained surrogate for stability-aware screening using one concrete materials example and one common failure mode.
• Applied anchor: constraint penalty ablation.
• Book anchor: [Neuer 6.1–6.3].

27. Case: explainability report for candidate shortlist

• Compare case: explainability report for candidate shortlist using one concrete materials example and one common failure mode.
• Applied anchor: model card template.
• Book anchor: [Neuer Ch7].

28. Case: post-mortem of failed discovery recommendation

• Diagnose case: post-mortem of failed discovery recommendation using one concrete materials example and one common failure mode.
• Applied anchor: sensitivity analysis report.
• Book anchor: [McClarren Ch11+12].

29. Balancing performance, interpretability, and cost

• Apply balancing performance, interpretability, and cost using one concrete materials example and one common failure mode.
• Applied anchor: failed recommendation post-mortem.
• Book anchor: [Murphy model selection].

30. Long-term maintenance of discovery models

• Define long-term maintenance of discovery models using one concrete materials example and one common failure mode.
• Applied anchor: deployment readiness checklist.
• Book anchor: [Bishop model limits].

31. Domain drift and model update policy

• Explain domain drift and model update policy using one concrete materials example and one common failure mode.
• Applied anchor: constraint penalty ablation.
• Book anchor: [Neuer 6.1–6.3].

32. Auditability and traceability requirements

• Compare auditability and traceability requirements using one concrete materials example and one common failure mode.
• Applied anchor: model card template.
• Book anchor: [Neuer Ch7].

33. Integration outlook with autonomous labs

• Diagnose integration outlook with autonomous labs using one concrete materials example and one common failure mode.
• Applied anchor: sensitivity analysis report.
• Book anchor: [McClarren Ch11+12].

34. Course-wide synthesis: from data to trusted decisions

• Apply course-wide synthesis: from data to trusted decisions using one concrete materials example and one common failure mode.
• Applied anchor: failed recommendation post-mortem.
• Book anchor: [Murphy model selection].

35. Capstone framing for students

• Define capstone framing for students using one concrete materials example and one common failure mode.
• Applied anchor: deployment readiness checklist.
• Book anchor: [Bishop model limits].

36. Exercise: constrained model mini-project

• Explain exercise: constrained model mini-project using one concrete materials example and one common failure mode.
• Applied anchor: constraint penalty ablation.
• Book anchor: [Neuer 6.1–6.3].

37. Exam checklist: trust argument with evidence

• Compare exam checklist: trust argument with evidence using one concrete materials example and one common failure mode.
• Applied anchor: model card template.
• Book anchor: [Neuer Ch7].

38. Advanced note: Physical Constraints, Trust, and Integration Outlook concept extension 33

• Diagnose advanced note: physical constraints, trust, and integration outlook concept extension 33 using one concrete materials example and one common failure mode.
• Applied anchor: sensitivity analysis report.
• Book anchor: [McClarren Ch11+12].

39. Advanced note: Physical Constraints, Trust, and Integration Outlook concept extension 34

• Apply advanced note: physical constraints, trust, and integration outlook concept extension 34 using one concrete materials example and one common failure mode.
• Applied anchor: failed recommendation post-mortem.
• Book anchor: [Murphy model selection].

40. Advanced note: Physical Constraints, Trust, and Integration Outlook concept extension 35

• Define advanced note: physical constraints, trust, and integration outlook concept extension 35 using one concrete materials example and one common failure mode.
• Applied anchor: deployment readiness checklist.
• Book anchor: [Bishop model limits].

41. Advanced note: Physical Constraints, Trust, and Integration Outlook concept extension 36

• Explain advanced note: physical constraints, trust, and integration outlook concept extension 36 using one concrete materials example and one common failure mode.
• Applied anchor: constraint penalty ablation.
• Book anchor: [Neuer 6.1–6.3].

42. Advanced note: Physical Constraints, Trust, and Integration Outlook concept extension 37

• Compare advanced note: physical constraints, trust, and integration outlook concept extension 37 using one concrete materials example and one common failure mode.
• Applied anchor: model card template.
• Book anchor: [Neuer Ch7].

43. Advanced note: Physical Constraints, Trust, and Integration Outlook concept extension 38

• Diagnose advanced note: physical constraints, trust, and integration outlook concept extension 38 using one concrete materials example and one common failure mode.
• Applied anchor: sensitivity analysis report.
• Book anchor: [McClarren Ch11+12].

44. Advanced note: Physical Constraints, Trust, and Integration Outlook concept extension 39

• Apply advanced note: physical constraints, trust, and integration outlook concept extension 39 using one concrete materials example and one common failure mode.
• Applied anchor: failed recommendation post-mortem.
• Book anchor: [Murphy model selection].

45. Exercise setup and dataset definition

• Define dataset, split protocol, and expected deliverables before any coding begins.
• Applied anchor: deployment readiness checklist.
• Book anchor: [Bishop model limits].

46. Exercise task 1 (pipeline core)

• Implement the core pipeline component with reproducible settings and documented assumptions.
• Applied anchor: constraint penalty ablation.
• Book anchor: [Neuer 6.1–6.3].

47. Exercise task 2 (comparison/ablation)

• Run an ablation/comparison under identical validation protocol and interpret differences.
• Applied anchor: model card template.
• Book anchor: [Neuer Ch7].

48. Exercise task 3 (failure analysis)

• Perform structured failure analysis and propose one evidence-backed mitigation.
• Applied anchor: sensitivity analysis report.
• Book anchor: [McClarren Ch11+12].

49. Exam-oriented key statements

• Summarize high-yield statements in concise written-exam style with definitions and caveats.
• Applied anchor: failed recommendation post-mortem.
• Book anchor: [Murphy model selection].

50. Summary, next-unit bridge, and references

• Consolidate the unit into a checklist: concepts, pitfalls, and decisions for next-unit transfer.
• Applied anchor: deployment readiness checklist.
• Book anchor: [Bishop model limits].