1bf8b2a684
docs: record EuRoC MH_01 real-run baseline across all doc surfaces
...
Updates README, testing/README, next_steps.md, and ADR 0001 with the
first real EuRoC MH_01 e2e run (100 frames, ~30s wall-time, ATE RMSE
~10.9 km → xfail). Places the EuRoC result alongside the prior VPAIR
baseline (~1770 km) so future-reader can see both failure modes at a
glance:
- VPAIR diverges because no raw IMU → ESKF never engages
- EuRoC diverges because indoor scene has no satellite anchor, so
VO+ESKF drift without an external correction
Also records the branching policy (rename ``euroc_mh01`` →
``euroc_machine_hall``; empty URL due to DSpace UI gate; manual
fetch via DOI 10.3929/ethz-b-000690084).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com >
2026-04-17 17:52:06 +03:00
03e617de63
docs(e2e): document VPAIR sample download + real-run status
...
Record first real e2e run on VPAIR sample (fixed-wing, 300-400 m
nadir): pipeline completes, ATE RMSE ~1770 km → xfail. VO without
IMU/satellite anchoring diverges on fixed-wing. Covered by xfail
branch; expected to flip to strict assert after VO+GPR tuning for
high-altitude nadir imagery.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com >
2026-04-16 23:24:41 +03:00
0062323d0a
docs(e2e): point README at dataset setup and test tiers
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com >
2026-04-16 22:03:14 +03:00
78dcf7b4e7
fix: post-audit — runtime bugs, functional gaps, docs, hardening
...
Phase A — Runtime bugs:
- SSE: add push_event() method to SSEEventStreamer (was missing, masked by mocks)
- MAVLink: satellites_visible=10 (was 0, triggers ArduPilot failsafe)
- MAVLink: horiz_accuracy=sqrt(P[0,0]+P[1,1]) per spec (was sqrt(avg))
- MAVLink: MEDIUM confidence → fix_type=3 per solution.md (was 2)
Phase B — Functional gaps:
- handle_user_fix() injects operator GPS into ESKF with noise=500m
- app.py uses create_vo_backend() factory (was hardcoded SequentialVO)
- ESKF: Mahalanobis gating on satellite updates (rejects outliers >5σ)
- ESKF: public accessors (position, quaternion, covariance, last_timestamp)
- Processor: no more private ESKF field access
Phase C — Documentation:
- README: correct API endpoints, CLI command, 40+ env vars documented
- Dockerfile: ENV prefixes match pydantic-settings (DB_, SATELLITE_, MAVLINK_)
- tech_stack.md marked ARCHIVED (contradicts solution.md)
Phase D — Hardening:
- JWT auth middleware (AUTH_ENABLED=false default, verify_token on /flights)
- TLS config env vars (AUTH_SSL_CERTFILE, AUTH_SSL_KEYFILE)
- SHA-256 tile manifest verification in SatelliteDataManager
- AuthConfig, ESKFSettings, MAVLinkConfig, SatelliteConfig in config.py
Also: conftest.py shared fixtures, download_tiles.py, convert_to_trt.py scripts,
config wiring into app.py lifespan, config-driven ESKF, calculate_precise_angle fix.
Tests: 196 passed / 8 skipped. Ruff clean.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com >
2026-04-02 18:27:35 +03:00
094895b21b
feat(phases 2-7): implement full GPS-denied navigation pipeline
...
Phase 2 — Visual Odometry:
- ORBVisualOdometry (dev/CI), CuVSLAMVisualOdometry (Jetson)
- TRTInferenceEngine (TensorRT FP16, conditional import)
- create_vo_backend() factory
Phase 3 — Satellite Matching + GPR:
- SatelliteDataManager: local z/x/y tiles, ESKF ±3σ tile selection
- GSD normalization (SAT-03), RANSAC inlier-ratio confidence (SAT-04)
- GlobalPlaceRecognition: Faiss index + numpy fallback
Phase 4 — MAVLink I/O:
- MAVLinkBridge: GPS_INPUT 15+ fields, IMU callback, 1Hz telemetry
- 3-consecutive-failure reloc request
- MockMAVConnection for CI
Phase 5 — Pipeline Wiring:
- ESKF wired into process_frame: VO update → satellite update
- CoordinateTransformer + SatelliteDataManager via DI
- MAVLink state push per frame (PIPE-07)
- Real pixel_to_gps via ray-ground projection (PIPE-06)
- GTSAM ISAM2 update when available (PIPE-03)
Phase 6 — Docker + CI:
- Multi-stage Dockerfile (python:3.11-slim)
- docker-compose.yml (dev), docker-compose.sitl.yml (ArduPilot SITL)
- GitHub Actions: ci.yml (lint+pytest+docker smoke), sitl.yml (nightly)
- tests/test_sitl_integration.py (8 tests, skip without SITL)
Phase 7 — Accuracy Validation:
- AccuracyBenchmark + SyntheticTrajectory
- AC-PERF-1: 80% within 50m ✅
- AC-PERF-2: 60% within 20m ✅
- AC-PERF-3: p95 latency < 400ms ✅
- AC-PERF-4: VO drift 1km < 100m ✅ (actual ~11m)
- scripts/benchmark_accuracy.py CLI
Tests: 195 passed / 8 skipped
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-04-02 17:00:41 +03:00
8649d13a78
docs: stage12 — Comprehensive README, full project documentation
2026-03-22 23:20:49 +02:00
766cef1514
feat: stage11 — Acceptance tests & performance benchmarks (80 tests)
2026-03-22 23:18:40 +02:00
c86cdc2e82
feat: stage10 — Full processing cycle with State Machine
2026-03-22 23:14:33 +02:00
74aa6454b8
feat: stage9 — Factor Graph and Chunks
2026-03-22 23:10:19 +02:00
905d6992de
feat: stage8 — Global Place Recognition and Metric Refinement
2026-03-22 23:03:54 +02:00
058ed315dd
feat: stage7 — Model Manager (F16) and Sequential VO (F07)
2026-03-22 22:59:55 +02:00
9ef046d623
feat: stage6 — Image Pipeline (F05) and Rotation Manager (F06)
2026-03-22 22:51:00 +02:00
a2fb9ab404
feat: stage5 — Satellite tiles (F04) and Coordinates (F13)
2026-03-22 22:44:12 +02:00
445f3bd099
feat: stage1 — domain schemas, SSE events, pydantic-settings config
2026-03-22 22:18:50 +02:00
Yuzviak