[JamesL787]

Summary

• `is_bosch_a_radar_car()` and its downstream handling in `radard.py` never landed alongside the Bosch-A radar decoder (from honda: port Bosch-A 16-slot radar decoder, enable for full HONDA_BOSCH_A harness #16, pending merge into nrdr/opendbc). Without it, decoded Bosch-A radar points feed into radard's generic path, which assumes the wrong Kalman filter timestep and has no real staleness/freshness gate for this radar's actual cadence.
• Ports the following (renamed to `HONDA_BOSCH_A` naming, not "civic_bosch"):
  • Kalman filter dt selected from the radar's real ~15Hz cadence (`HONDA_BOSCH_A_RADAR_TS`), not the ~20Hz model-cycle default
  • `measurement_update` gating so a duplicate `radarTracks` payload (radar runs slower than the model loop) isn't absorbed twice by the KF
  • a "preferred track" continuity path: a previously-selected lead stays selected through ordinary model-probability fluctuations as long as it's still coherent, instead of re-arbitrating from scratch every cycle
  • staleness detection with two independent triggers (a better-scoring challenger track sustained over several cycles, or gross absolute-range disagreement with the model sustained over several cycles) that drops a stale preferred track rather than holding it indefinitely
  • `honda_bosch_a_low_speed_radar_lead_sane()`, requiring a few real sweeps before a radar-only low-speed takeover
• Purely additive relative to this file's current content -- all other car behavior is untouched.
• Mechanical, unmodified transplant of behavior already tested on the branch this was developed on (109/109 relevant tests passing there) onto this repo's current `radard.py`.

Test plan

• Ported unmodified from a branch where it passes 109/109 relevant tests
• Run through this repo's own radard/Honda test suite before merge
• On-road validation alongside the Bosch-A decoder once honda: port Bosch-A 16-slot radar decoder, enable for full HONDA_BOSCH_A harness #16 is merged

🤖 Generated with Claude Code