Scenario context
Warehouse AGVs must plan paths in dynamic aisles, yield to peers under traffic rules, dock precisely with racks, and recover from faults without unsafe motion. Harmony-SE excels at multi-party interactions among WMS, fleet dispatch, onboard controllers, and floor staff.
The reference scenario is synthetic: no real warehouse operator, AGV OEM, or customer SLAs are named. It still reflects production pain — deadlock near choke points, last-centimeter docking, lost Wi-Fi segments, and handoff to manual jack mode.
Actors and contracts
WMS publishes pick missions; dispatch assigns routes and priorities; the vehicle controller executes trajectories and reports telemetry; humans intervene during exceptions. Harmony-SE use cases make those contracts explicit before engineers dive into state machines.
States and interactions
Typical states include idle, navigate-to-pick, fine docking, fault hold, and manual takeover. The KB slices content by state machine and sequence diagrams so AI4MBSE can focus one diagram at a time — for example modeling yield behavior at a narrow intersection without rewriting the entire fleet model.
Scheduling and safety
Dispatch policies encode priority inversions for emergency picks, speed caps in pedestrian zones, and geofenced maintenance bays. Conversational modeling should attach numeric policies to elements reviewers can inspect, aligning with human-in-the-loop write-back rather than opaque LLM prose.
Tooling path
Teams may model in Capella for logical architecture, SysON for SysML v2 experiments, or commercial SysML suites. Ark kernel plus adapter strategy keeps assistant behavior consistent across hosts while respecting each tool project structure.
Telemetry and replay
Fleet operators replay incident logs when AGVs deadlock or miss dock targets. The reference includes trace hooks from telemetry events to state-machine transitions so KB retrieval during troubleshooting sessions surfaces the same vocabulary dispatchers use — supporting audit-friendly conversational modeling.
MBSE RAG value in this scenario
Industry reference scenarios exist so the engineering knowledge base can store reusable slices — requirements phrasing, allowed diagram sequences, naming prefixes, and verification hooks — without exposing any real customer program. When AI4MBSE runs a one-diagram-at-a-time session, MBSE RAG should pull only the fragments that belong to the active methodology and view. That keeps conversational modeling honest: candidates stay checkable, human-in-the-loop write-back stays authoritative, and reviewers still own baselines under configuration management aligned with INCOSE practice and OMG SysML / SysML v2 vocabulary.