Ingot documentation

Core value: Turn every real recipe run into optimization evidence and continuously recommend the next recipe within safety boundaries and observed coverage.

Product category: Open-source Process Diagnosis & Optimization. Ingot links run conditions, process trajectories, production context, and quality outcomes so normal recipe runs become optimization observations that support the next recipe recommendation.

The documentation can be read by task rather than in sequence.

Choose documentation by task

Product loop

Process configuration → Field integration → Production runs → Quality management → Process diagnosis → Recipe optimization
           ↑                                                                                         ↓
           └──────── Validated specifications, operating regions, and knowledge return to production ────────┘

1. Process configuration tells the system which variables, units, quality rules, and safety boundaries matter.
2. Field integration turns control, instrument, and business data into consistent process fields.
3. Production runs record actual conditions, stages, trajectories, and manufacturing context.
4. Quality management links inspections uniquely and subjects them to independent review.
5. Process diagnosis checks whether the data are reliable, compares runs, and finds differences worth testing.
6. Recipe optimization aggregates real recipe runs and recommends the next recipe within safety boundaries and observed coverage; controlled validation is optional when causal confirmation or extrapolation is needed.

This order means “what must exist before the next step.” Navigation may follow day-to-day role needs, but analysis must still begin with trustworthy data.

Current maturity

The main software workflow runs and is covered by automated tests. The repository bundles no scenario-specific validation data or results; users evaluate optimization effects, applicability, and production safety with their own data.

See Current status for the complete boundary.

Documentation map

Development and operation

• Getting started: synthetic tour and complete local stack
• Recipe-optimization pilot guide: move from real runs to the first next-recipe recommendation
• Data integration: identity, protocols, points, mappings, and data admission
• Deployment: configuration, health, monitoring, backup, and upgrade
• Frequently asked questions: concise answers on product boundaries and method choice

System and algorithm design

• System design: stable business model and component responsibilities
• Analysis and optimization: how real recipe runs become observations and support the next recipe recommendation
• Mechanism knowledge design: sources, claims, review, fusion, and applicability

Validation and production engineering

• Current status: what works today and what remains unproven
• Scenario validation: how historical review, shadow use, and controlled experiments test value
• Production architecture: what a production deployment must satisfy

Project governance

• Roadmap: long-term direction, priorities, and promotion gates
• Brand guide: product positioning and public wording boundaries
• Open-source dependencies: introduction and audit principles

Reading status labels

• Current operating guide: procedures that can be executed with the current release.
• Current facts: what is implemented, validated, and still limited.
• Architecture or specification baseline: boundaries the product must continue to respect.
• Staged implementation or target design: includes future work and is not current behavior.
• Rolling roadmap: build order changes as evidence changes.

Numeric facts are not copied across entry documents. Brand guide governs the product category, Current status governs maturity, and the public validation record governs method figures.

中文文档从 index.md 开始。