Module 12 · Phase 4: Localization and autonomy · Flagship
Outdoor-to-Indoor Localization & GNSS-Denied Navigation
Flagship module. A robot leaves GNSS coverage and enters a building with no pose jump. This is a covariance-aware estimator handover with aligned frames and a recovery plan, not a SLAM on/off switch.
This overview is open. 6 of 6 guides are written; the others show Coming soon until they are complete and tested.
Questions this module answers
- How do I detect GNSS degradation and decide when to stop trusting it?
- How do I initialize indoor SLAM and align indoor and outdoor maps?
- How do I switch localization sources without a pose jump and keep global continuity?
- What happens if LiDAR localization fails, and how do I recover?
What it covers
- GNSS quality monitoring and degradation detection
- Frame definitions and map alignment
- Continuous global pose, update-rate differences
- Covariance-aware handover and confidence monitoring
- Relocalization and map consistency
- Failure scenarios and validation
Guides and design decisions
A. Engineering Guides (5)
B. Design & Selection (1)
Diagnostics, tools and challenges
C. Troubleshooting (5)
- GNSS-to-indoor transition causes a pose jumpThe pose steps when GNSS is dropped near a building.
- LiDAR localization becomes unstableThe pose oscillates or snaps between hypotheses.
- Localization covariance grows unexpectedlyReported uncertainty rises with no obvious sensor loss.
- Map and odometry frames are inconsistentThe map to odom transform jumps or fights with another publisher.
- Localization fails after rebootThe robot cannot find its position after a restart.
D. Interactive tools (1)
- Localization Handover VisualizerOutdoor-to-indoor transition: GNSS, LiDAR and VIO confidence, pose continuity and covariance evolution.
E. Design challenges (1)
- Challenge 2: Outdoor-to-Indoor Autonomous NavigationGNSS outdoors to LiDAR-inertial or visual-inertial indoors: sensor and frame architecture, sync, calibration, estimation, handover and recovery.
Practical deliverables
- Architecture diagrams and data flow
- Frame definitions
- Example ROS 2 configurations
- Failure-scenario table
- Validation procedure
Read first: existing FixTheRobot lessons and labs
This module builds on these instead of repeating them.
- GPS-Denied Localization SLAM · Lesson
- Mapping vs Localization Mode SLAM · Lesson
- Map Saving and Reuse SLAM · Lesson