ROS 2 Kinematic Guard adds fleet telemetry for VDA5050 command integrity
Wi-Fi collapse stress test shows fleet systems can hold new orders until vehicle recovers.
The updated ros2_kinematic_guard project bridges a missing layer between vehicle-internal command execution and fleet-level observability. Originally a local /cmd_vel guard, it now translates NARH Guard state into structured telemetry consumable by ROS diagnostics, MCAP post-analysis, and VDA5050-style orchestration layers. During a Wi-Fi collapse stress test, it outputs states like 'RESYNCING' with latency=CRITICAL and R_NAR values (e.g., 5.478, 1412.078), along with vehicle_response='RESYNC_REQUIRED' and fleet_action='HOLD_NEW_ORDERS'. This allows fleet managers to avoid assigning intersection-heavy or timing-critical maneuvers until the vehicle recovers to 'RECOVERED' state.
Unlike simple timeouts that only detect silence, the NARH Guard detects silence plus local predictive braking, stale commands via timing and kinematic drift, burst commands via residual spikes, and replay/out-of-order via phase-continuity checks. It also measures command/odometry consistency directly. The recovery logic is state-aware with a resync gate, and fleet visibility extends beyond internal failure to structured telemetry (ROS/VDA5050). This is not a replacement for ros2_control controller-side mechanisms but a higher-level signal for mixed-fleet integration and post-incident analysis.
- Detects silence plus local predictive braking, not just timeout gaps
- Exposes structured telemetry via ROS diagnostics and VDA5050-style topics
- State-aware resync gate with recovery logic prevents new orders during degraded mode
Why It Matters
Mixed-fleet orchestrators can now quantitatively trust command integrity, improving safety in degraded network conditions.