armteam/phantom-episodes

PHANTOM episodes

Tactile manipulation episodes for the PHANTOM tactile world-action-model (UR3 + Robotiq 2F-85 + 2x Daimon DM-Tac W2L fingertip sensors + RealSense scene camera, teleoperated via the Echo exoskeleton leader).

Layout

• tasks/<task>/ep_<task>_<epoch>_<idx>/ — canonical flattened view: one directory per episode, grouped by task. Use this + manifests/.
• manifests/<task>.jsonl, manifests/all.jsonl — one JSON object per episode: success, split (train/val), tactile_contact, duration, peak force, contact area, source session, quality notes, hub path.
• manifests/quality_full.csv — full per-episode quality audit.
• archive/, collect/ — raw as-recorded session layout (provenance; superset that also contains pre-cleanup debug takes).

Episode format

Each episode directory holds one zarr group per stream (data + ts arrays, timestamps in master-clock seconds) plus meta.json. Streams:

Stream	Contents	Rate
camera_scene_color	JPEG images	~15 Hz
tactile_left, tactile_right	wrench (6D, N / 1e−2 Nm), area (mm²), fields_ds (72×96×8, f16), keyframes (144×192×8, f16), infer_img (288×384, u8)	~6.5 Hz
arm	q, qd, tcp_pose, tcp_speed, ft	~125–140 Hz
gripper	pos, obj-detect	~100 Hz
actions	Delta end-effector commands	~10 Hz
actions_abs	Absolute joint targets + gripper command	Event-based

Tasks

task	episodes	role	median dur	median peak |F|	median contact
Carton	180	successes	17.4 s	12.9 N	17.9 mm²
Carton_fail	20	failure demos	17.7 s	26.3 N	36.1 mm²
waffles	180	successes	18.3 s	8.4 N	3.0 mm²
waffles_fail	20	failure demos	20.4 s	17.8 N	25.5 mm²
egg	155	successes	27.6 s	8.8 N	18.0 mm²
egg_fail	20	failure demos	18.4 s	33.5 N	49.1 mm²

Notes

• success=true everywhere in the success tasks; *_fail tasks are DELIBERATE failure demonstrations (recorded as intended, higher forces/contact — contrastive signal).
• tactile_contact=false marks episodes where the grasp landed outside the sensor pads (valid vision/proprio demos, no tactile signal).
• Split rule: per success task, the last 2 sessions (chronological) are val, the rest train; failure demos are train-only. Re-split freely via the manifests — they are the source of truth, not the folder layout.
• Peak forces briefly exceed the 30 N pad ceiling in a handful of episodes (dynamic spikes, mostly failure demos) — flagged in quality_notes.