* Remove `webots_quick` Makefile target — `make webots` is the only
webots entry point now (it fires the interactive picker). The
positional non-interactive path is still available as
`bash tools/run_webots.sh N MODE DRIVE WORLD` for scripted use.
* Add `WEBOTS_PRESET_FLAG = --use-webots-preset` for mecanum drive
and pass it to the `bc.collect` recipe so demos are collected
under the gym kinematics that match the physical-roller Webots
mecanum. Without this, mecanum BC demos would record textbook
X-pattern teacher actions against textbook gym kinematics, and
the resulting policy would fail at deployment exactly the same
way the current v1 mecanum policies do.
* `rl/train.py` already auto-detects mecanum and applies
HERDING_MEC_WEBOTS internally (commit 3b4c99a), so the rl recipe
doesn't need the flag — a one-line comment in the Makefile makes
that intent explicit.
Diff drive keeps the existing recipe: no --use-webots-preset, so
BC demos collected on HERDING_DEFAULT (360° gym, no FP). This is
the regime that produced the current diff/field and diff/round
policies that pen 5/5 in Webots LiDAR; retraining under the same
regime is the safest reproduction.
126 pytest cases still pass.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* `controllers/shepherd_dog/shepherd_dog.py`
- Tracks the first step at which each sheep crosses the gate; on
auto-finish (all sheep penned) prints a `[results]` summary
block: mode/drive/world/lidar/dogs/seed line, total simulated
time, per-sheep penning order with absolute step + seconds since
sim start, and the gate spread between the first and last
penning.
- Reads `HERDING_SEED` (env / runtime cfg) and seeds the
controller's RNG when set. Empty = time-based default = old
non-deterministic behaviour.
* `controllers/sheep/sheep.py` reads `HERDING_SEED` the same way
(loading `herding_runtime.cfg` itself so it works even when
Webots strips env vars) and seeds Python's RNG XOR'd with the
sheep's name hash, so a fixed seed gives a reproducible flock
trajectory without all sheep starting from identical wander state.
* `tools/run_webots.sh` writes `HERDING_SEED` into the runtime cfg
(empty when unset so existing scripts keep their stochastic
behaviour).
* `tools/webots_menu.sh` gains a Seed prompt (random / fixed
integer); the launch summary box shows the choice next to the
perception row.
* `Makefile`
- `make webots` now fires the interactive picker (replacing the
old positional invocation).
- `make webots_quick MODE=… DRIVE=… WORLD=… N=…` is the old
positional path, kept for batch / scripted use.
Smoke-tested: menu renders Mode → Drive → World → LiDAR → Dogs
→ Sheep → Perception → Seed → Headless prompts and shows the
selected Seed value in the launch summary. 126 pytest cases still
pass.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Today's session worked across the full Webots delivery stack — found and
fixed a cluster of bugs blocking the BC/RL transfer, then explored
training-side mitigations for the residual perception gap.
Bug fixes:
- Makefile FP_RATE default 2.0 → 0.0: BC demos used fp_rate=0 but RL
fine-tune defaulted to fp_rate=2, poisoning the BC obs distribution
and stalling PPO at 0% success across 1.46M+ steps.
- controllers/{shepherd_dog,sheep}/runtime.ini: Webots was launching
controllers under system python3 (no numpy) and they were crashing
silently. Pinned to the conda tir env.
- herding/config.py HERDING_WEBOTS preset: pen_latch_depth 0.5 → 2.0,
max_new_tracks_per_step 3 → 1, static_reject 0.8 → 1.2. Stops phantom
FPs near the gate from latching as permanently-penned tracks.
- herding/perception/sheep_tracker.py: penned tracks now decay at
forget_steps × 8 instead of living forever. Adds get_positions
min_freshness filter for deploy-time use.
Training/eval matches deployment:
- training/bc/collect.py: --dagger-policy flag for DAgger rollouts
(policy drives, teacher labels) + --use-webots-preset for matched
140° tracker + DR config.
- controllers/shepherd_dog/shepherd_dog.py: scan-fallback (0, 0.6) when
BC/RL sees empty sheep_positions — recovers from FOV gaps.
Tooling:
- tools/dagger_round.sh: one-shot DAgger round (collect + concat + bc).
- tools/webots_sweep_gt.sh: full sweep with HERDING_USE_GT=1 for the
perception-gap diagnosis matrix.
- protos/ShepherdDog360.proto: 360° FOV variant for the FOV-ablation
comparison. Canonical proto stays at 140° per project spec.
Artifacts: v1 BC/RL policies for all 4 (drive × world) combos trained
in clean gym (success: diff/field 90-100%, diff/round 58%, mec/field
60-100%, mec/round 50-100%). DAgger r1/r2 BCs for diff/field show
12%→38% progression on gym HERDING_WEBOTS proxy but did not close
to actual Webots LiDAR (0/5 throughout). Next: LSTM policy or
learned tracker per the project-state memory.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Johnny Fernandes