docs: align roadmap with tender-fit requirements

Plan aircraft-risk modelling, CCSDS RDM support, tender-grade replay validation, and ESA software assurance artefacts in the implementation and master plans.
This commit is contained in:
Allissa Auld
2026-04-17 20:31:37 +02:00
parent 2c49d7cd5c
commit 266ee30d4b
2 changed files with 15819 additions and 0 deletions

303
docs/Implementation_Plan.md Normal file
View File

@@ -0,0 +1,303 @@
# SpaceCom Implementation Plan
## Purpose
This document turns the master plan into an implementation-oriented coding plan. It is focused on phases of engineering work and sprint-style delivery, not on full architectural detail.
## Delivery Model
The implementation plan assumes:
- self-hosted GitLab as the delivery platform
- Docker Compose as the baseline deployment model
- contracts as the authoritative entitlement source
- Phase 0 blockers cleared before major coding commitments
- safety-critical, auth, and regulatory-sensitive changes always require human review
## Phase 0: Pre-Build Decisions
Objective:
- clear the decisions that would otherwise force rework
Sprint 0 outcomes:
- Space-Track AUP architecture decision recorded
- Cesium commercial licence executed
- GitLab CI/CD authority confirmed
- contract-driven entitlement model confirmed
- Redis trust split approved
- initial ADR set aligned with these decisions
Exit criteria:
- no unresolved blocker remains that would change ingest, licensing, or deployment direction
## Phase 1: Foundation Sprints
Objective:
- establish the platform baseline and developer workflow
### Sprint 1.1
- backend project structure and FastAPI scaffolding
- frontend project structure and Next.js scaffolding
- Docker Compose topology
- TimescaleDB/PostGIS baseline schema
- GitLab CI skeleton
- docs tree, ADR format, `AGENTS.md`, changelog baseline
### Sprint 1.2
- auth baseline: JWT, cookies, MFA foundation
- RBAC and role checks
- liveness and readiness endpoints
- structured logging
- security scan hooks and pre-commit
- first-login acceptance gates for ToS/AUP/Privacy Notice
Exit criteria:
- full stack starts cleanly
- CI runs on every merge request
- auth and role boundaries exist
- baseline docs and controls are in place
## Phase 2: Data and Ingest Sprints
Objective:
- make the platform ingest and manage real operational data safely
### Sprint 2.1
- object catalog CRUD
- TLE ingest
- ingest worker wiring
- source validation and checksum checks
- basic data freshness indicators
### Sprint 2.2
- DISCOS import
- space weather ingest
- retention policy scaffolding
- ingest metrics and failure alerts
- first degraded-mode backend signals
Exit criteria:
- core upstream data can be ingested, validated, stored, and monitored
## Phase 3: Propagation and Physics Core
Objective:
- deliver the first real analysis engine
### Sprint 3.1
- frame/time utilities
- SGP4 propagation
- frame transformation validation
- initial CZML generation
### Sprint 3.2
- numerical decay predictor
- atmospheric inputs
- upper/lower atmosphere model interfaces for terminal descent
- model versioning hooks
- prediction input validation
Exit criteria:
- the platform can produce technically valid propagated and decay outputs with traceable inputs
## Phase 4: Simulation and Hazard Engine
Objective:
- build the uncertainty and event-analysis layer
### Sprint 4.1
- Monte Carlo job orchestration
- Celery chord flow
- worker resource controls
- result aggregation guards
### Sprint 4.2
- breakup logic and sub/trans-sonic fragment model baseline
- fragment shape, tumble, and descent-assumption reference data
- corridor generation
- object/event linking
- first event orchestration path
Exit criteria:
- multi-run hazard outputs are generated safely, bounded, and auditable
## Phase 5: Operational Frontend
Objective:
- make the product usable for analysts and operators
### Sprint 5.1
- globe and layer controls
- timeline strip
- object/event navigation
- degraded-state presentation
### Sprint 5.2
- alert centre
- acknowledgement flow
- decision prompts
- keyboard-first interaction paths
Exit criteria:
- operators can see, interpret, and act on platform outputs in the UI
## Phase 6: Event Workflow and Reporting
Objective:
- support full event handling, not just display
### Sprint 6.1
- event detail page
- event state transitions
- report job submission and tracking
- WebSocket event flow
### Sprint 6.2
- report rendering pipeline
- export storage flow
- report audit trail
- offline/reconnect handling
Exit criteria:
- an end-to-end event can be analysed, reviewed, and reported through the platform
## Phase 7: Reliability, Safety, and Compliance Hardening
Objective:
- turn a capable system into an operationally defensible one
### Sprint 7.1
- observability dashboards
- SLO instrumentation
- alerting rules
- incident-response runbook alignment
### Sprint 7.2
- privacy workflows
- retention and pseudonymisation tasks
- safety-case artefacts linkage
- audit-log separation and integrity controls
Exit criteria:
- the system is observable, auditable, and aligned with documented operational obligations
## Phase 8: Aviation Product Expansion
Objective:
- strengthen the ANSP-facing differentiation
### Sprint 8.1
- FIR intersection and aviation-context outputs
- aircraft exposure and vulnerability scoring
- uncertainty display refinements
- operator-facing airspace impact summaries
- conservative-baseline comparison outputs
### Sprint 8.2
- shadow mode behaviour
- multi-ANSP coordination foundations
- NOTAM-adjacent workflow support
- live air-traffic density and route inputs for air-risk products
Exit criteria:
- the aviation-side product layer is credible for shadow deployment and stakeholder review
## Phase 9: Space Operator Expansion
Objective:
- extend the shared core into upstream customer workflows
### Sprint 9.1
- owned-object data model and access rules
- API key lifecycle
- portal foundations
### Sprint 9.2
- controlled re-entry planner
- CCSDS OEM/CDM/RDM export
- partner-facing integration workflows
Exit criteria:
- the shared physics core now supports both downstream aviation and upstream space-operator use
## Phase 10: Deployment Readiness and Evidence
Objective:
- prepare for shadow deployment and external scrutiny
### Sprint 10.1
- validation and backcasting workflow
- named historical replay corpus and conservative-baseline comparisons
- tender-grade KPI and air-risk validation pack
- shadow reporting outputs
- deployment gating refinements
### Sprint 10.2
- final compliance artefacts
- ESA software assurance artefacts: SRS, SDD, user manual, test report
- release checklist completion
- operational readiness review
- first shadow deployment package
Exit criteria:
- the platform is ready for controlled external deployment and evidence-based review
## Cross-Cutting Workstreams
The following run continuously across phases rather than in only one sprint:
- security architecture and dependency hygiene
- documentation and ADR maintenance
- test coverage and validation artefacts
- safety-case traceability
- legal/commercial gating
- performance and capacity reviews
- procurement traceability and bid artefact maintenance
## Recommended Execution Pattern
For actual delivery, each sprint should end with:
- a demoable increment
- updated docs
- updated ADRs where needed
- passing CI
- explicit review of safety/compliance impacts
## Relationship to Other Docs
- [Overview.md](/d:/Projects/SpaceCom/docs/Overview.md) gives the shortest summary
- [Roadmap.md](/d:/Projects/SpaceCom/docs/Roadmap.md) gives the big-picture staged view
- [MASTER_PLAN.md](/d:/Projects/SpaceCom/docs/MASTER_PLAN.md) remains the authoritative detailed plan