AI Augmented Systems Architect — Industry

Industry ADRs that AI agents can resume from; safety functions with SIL or PL, verification and ALARP; your plant's IEC 62443 zones and conduits (not bRRAIn's platform zones); discipline overlays; brownfield zoning from discovered flows; and a request that crosses safety and security boundaries.

1. Pretest: industry ADRs, safety zones and IEC 62443 Diagnostic pretest · 5 min
2. Industry ADR vs software ADR — different cadence, different stakes Reading · 15 min
3. Safety zones — SIL ratings, ALARP, and how AI must reason about them Reading · 15 min
4. Security zones and conduits — IEC 62443 mapped to your plant Reading · 15 min
5. Discipline overlay — zones in Construction, Mining, Manufacturing and Civil Reading · 15 min
6. Worked example — zoning a brownfield concentrator from discovered flows Worked example · 15 min
7. Scenario — the integrator wants a write path into the safety PLC Scenario · 15 min
8. Lab 1: Industry ADRs and a safety-and-security zone map AI role-play lab · 45 min
9. Retrieval: 8 questions across the module Retrieval check · 10 min

Asset Administration Shells, OPC UA companion specifications, IEC 61850 SCL and ISA-95 placement; turning P&IDs and data sheets into verified knowledge with Mega Parser and the vault; and a worked example keeping five artefacts consistent with one set of IDs.

1. Pretest: machine-readable industrial architecture Diagnostic pretest · 5 min
2. AAS anatomy — shells, submodels and the asset's canonical description Reading · 15 min
3. OPC UA companion specifications — how an industry agrees on an information model Reading · 15 min
4. IEC 61850 SCL — substation configuration as a graph an agent can traverse Reading · 15 min
5. ISA-95 levels — why the level 2/3 and 3/4 boundaries are where AI architecture lives Reading · 15 min
6. From P&IDs and data sheets to governed knowledge — Mega Parser and the vault Reading · 15 min
7. Worked example — one conveyor, five artefacts, one set of IDs Worked example · 15 min
8. Lab 2: Render one system as consistent machine-readable artefacts AI role-play lab · 45 min
9. Retrieval: 8 questions across the module Retrieval check · 10 min

Safety and reliability requirements as checkable preconditions; availability, recovery, timing and data freshness as invariants; the 'AI says no' pattern; resolving conflicts between invariants with different owners; and a worked invariant checker wired into CI.

1. Pretest: constraints and NFRs as invariants Diagnostic pretest · 5 min
2. Safety and reliability NFRs as invariants — SIL, proof testing, MTBF, MTTR Reading · 15 min
3. Availability and latency as invariants — RPO, RTO, scan time and data freshness Reading · 15 min
4. The 'AI says no' pattern — refusals that cite the invariant and show the path Reading · 15 min
5. Scenario — when the availability target and the proof-test interval collide Scenario · 15 min
6. Worked example — an invariant checker and CI gate for an architecture set Worked example · 15 min
7. Lab 3: Six NFRs as invariants, and the refusal that cites them AI role-play lab · 45 min
8. Retrieval: 8 questions across the module Retrieval check · 10 min

Evolution paths for capacity, conversion, retrofit and decommissioning; drift budgets; drift detection you implement across topology, telemetry, control and documentation; digital-twin alignment; refactor ADRs; and triage of a batch of drift findings on a bridge monitoring system.

1. Pretest: evolution, drift and the digital twin Diagnostic pretest · 5 min
2. Evolution paths — capacity, conversion, retrofit and decommissioning Reading · 15 min
3. Drift detection — topology, telemetry, control and documentation Reading · 15 min
4. Digital twin alignment — keeping the twin, the records and reality in agreement Reading · 15 min
5. The refactor ADR — deciding which way to align, and proving it Reading · 15 min
6. Scenario — five drift findings on the bridge monitoring system Scenario · 15 min
7. Lab 4: Triage a drift report and write the refactor ADR AI role-play lab · 45 min
8. Retrieval: 8 questions across the module Retrieval check · 10 min

The cost, availability, safety and latency matrix with a precise safety override; discipline pattern catalogs; AI and human ownership for each step of a decision; and a worked triage of a construction crane-data request.

1. Pretest: trade-offs, patterns and ownership Diagnostic pretest · 5 min
2. The cost / availability / safety / latency matrix — weighing plant changes explicitly Reading · 15 min
3. Discipline pattern catalogs — matching new requests to proven solutions Reading · 15 min
4. AI and human ownership by axis — who does which step of a decision Reading · 15 min
5. Worked example — triaging a crane-data request on a construction site Worked example · 15 min
6. Lab 5: Triage five modification requests for a plant leadership meeting AI role-play lab · 45 min
7. Retrieval: 8 questions across the module Retrieval check · 10 min

AI-augmented HAZOP in which the team rules; curation of the living architecture set and its place in the 9-stage Build Methodology; architecture memory with the bRRAIn session tools, vault layout and per-project permissions; the knowledge graph and Graph Editor; and compliance mapping with Robo Compliance.

1. Pretest: reviews, living practice and architecture memory Diagnostic pretest · 5 min
2. AI-augmented HAZOP — the agent prepares the deviations, the team rules Reading · 15 min
3. Living-set hygiene — the architect as curator Reading · 15 min
4. Architecture memory in bRRAIn — the session method, the vault and who can see what Reading · 15 min
5. System maps in the knowledge graph, and compliance mapping with Robo Compliance Reading · 15 min
6. Lab 6: Run an AI-augmented HAZOP and land the outcome in memory AI role-play lab · 45 min
7. Retrieval: 8 questions across the module Retrieval check · 10 min

The four discipline capstone briefs, the architecture pack and the review format; preparing the pack in dependency order; retrospectives with honest AI and human attribution; the four anchor exemplars; and a calibration lab scoring an unfamiliar capstone.

1. Pretest: capstone preparation Diagnostic pretest · 5 min
2. Capstone brief — what you receive, what you prepare, how the review runs Reading · 15 min
3. Preparing the architecture pack — sequence, checkpoints and using your assistant well Reading · 15 min
4. Retrospectives with honest attribution — what the AI did, what the humans did Reading · 15 min
5. Anchor exemplars — what 92, 78, 71 and 54 look like Reading · 15 min
6. Lab 7: Score an unfamiliar capstone against the rubric AI role-play lab · 45 min
7. Retrieval: 7 questions across the module Retrieval check · 10 min

The exam blueprint, interleaved review across all domains, timing strategy, the seven common miss patterns, timed performance-task practice and the capstone design review.

1. Pretest: exam readiness Diagnostic pretest · 5 min
2. The exam blueprint — what is tested, how much, and in what form Reading · 15 min
3. Interleaved review — one site, every domain Scenario · 15 min
4. Timing strategy — two passes, flags, and protecting the performance tasks Reading · 15 min
5. Seven common miss patterns — and the move that fixes each Reading · 15 min
6. Lab 8: Two timed performance tasks with examiner follow-ups AI role-play lab · 45 min
7. Retrieval: 7 questions across the course Retrieval check · 10 min
8. Capstone: defend your architecture pack in a design review AI role-play lab · 45 min