armteam/hapticwam-rig-episodes

HapticWAM — closed-loop rig episodes

HapticWAM: Distilling Imagined Touch into a World-Action Model without Inference-Time Tactile Sensing — paper arXiv:2609.23888, submitted to ICRA 2027. Code: github.com/Advanced-Robotic-Manipulation/HapticWAM · all repos: HapticWAM — ICRA 2027.

Recorded rollouts from the physical rig for HapticWAM (Haptic World-Action Model), the tactile world-action model built on Cosmos-Predict2.5-2B: a tactile-conditioned teacher whose imagined contact future is distilled into a pad-free student. Rig: UR3 CB3 over RTDE + Robotiq 2F-85 with Daimon DM-Tac W2L fingertip pads + RealSense scene camera. These are the deployment takes behind the 2026-09-16 closed-loop evaluation (teacher, pad-free student, pi0.5 and Diffusion Policy baselines).

Renamed from armteam/phantom-rig-20260915 on 2026-09-19, when the project's working name PHANTOM became HapticWAM. The old id still redirects. The Python package and CLI keep the name phantom, so checkpoint names, config keys and run names are unchanged.

Layout

Path	What
20260915_experiment/	the analysis set — the takes the reported numbers are computed from
20260915_experiment_extra/	additional takes recorded in the same sessions, outside the analysis set
20260915_experiment_superseded/	takes replaced by a later re-run; kept for provenance

One directory per episode, named ep_<arm>_<task>_<epoch>_<idx>, where <arm> identifies the policy under test. Tactile is recorded in every mode, including the pad-free student runs, so the student's takes can be scored against contact it never saw.

Episode format

Each episode directory holds one zarr group per stream (a data array plus a ts array of master-clock timestamps) and a meta.json: camera_scene_color, tactile_left, tactile_right, arm (q, qd, tcp_pose, tcp_speed, ft), gripper, actions, actions_abs. The same format as armteam/hapticwam-teleop-raw, which documents the streams and rates in full.

Schema

One directory per episode: meta.json plus one zarr group per stream, each holding a data array (T, …) and a ts array (T,) of master-clock seconds. Every stream keeps its own T and its own timestamps — nothing is resampled onto a common grid — so align by nearest ts (observations) or first ts >= t (future targets).

Stream	Shape	dtype	Rate	Meaning
tactile_{left,right}_fields_ds	(T, 72, 96, 8)	f16	~5.7 Hz	tactile field stack, channels [disp_x, disp_y, depth, shear_x, shear_y, fx, fy, fz]; depth and the distributed force fx,fy,fz are in raw SDK units (uncalibrated)
tactile_{left,right}_keyframes	(T, 144, 192, 8)	f16	~2.5 Hz	the same 8 channels at higher spatial resolution, time-decimated
tactile_{left,right}_infer_img	(T, 288, 384)	u8	~5.7 Hz	the SDK's gel image (getInferImg)
tactile_{left,right}_wrench	(T, 6)	f32	~5.7 Hz	pad wrench [Fx, Fy, Fz, Mx, My, Mz], N and N·m (SI: the SDK's 1e−2 N·m torques are scaled at the driver boundary)
tactile_{left,right}_area	(T,)	f32	~5.7 Hz	contact area, mm²
arm_q, arm_qd	(T, 6)	f64	125 Hz	joint position (rad), joint velocity (rad/s)
arm_tcp_pose, arm_tcp_speed	(T, 6)	f64	125 Hz	TCP pose [x, y, z, rx, ry, rz] (m + rotation vector, base frame) and twist
arm_ft	(T, 6)	f64	125 Hz	wrist wrench, N / N·m — on this CB3 arm a current-based estimate with a large pose-dependent bias, not a real F/T sensor
gripper	(T, 2)	f32	~95 Hz	[position, obj]: closure 0 (open) … 1 (closed), and the Robotiq gOBJ status 0..3 (2 = stopped by contact while closing, i.e. holding)
camera_scene_color	(T, 480, 640, 3)	u8	15 Hz	scene RGB, JPEG-encoded per frame (quality 92) in a zarr VLenBytes object array; the group's attrs carry encoding, frame_shape, frame_dtype, jpeg_quality
actions	(T, 7)	f32	10 Hz	the canonical action: Δ-EE pose step [Δx, Δy, Δz, Δrx, Δry, Δrz] (m, rotation-vector rad) + commanded gripper closure
actions_abs	(T, 7)	f32	10 Hz	absolute command [q_target(6) rad, gripper]

Rates are what the rig actually achieved (the tactile SDK free-runs below its 8 Hz cap); use ts, never an assumed rate. A stream that produced no samples has no directory at all, so check before you read: a pad-free deploy (tag padfree:on) has no tactile_* streams, simulation exports and most deploy takes have no actions_abs (and sim adds contact_gt), and a re-derived policy rollout adds actions_plan (the executor's pre-clamp proposal).

meta.json carries task, text, operator, tags, policy, dagger_round, success, damage, notes, driver_modes, clock_calibration, config_hash, hardware_shapes, deploy_overrides, status and weight. It is authoritative for task and labels — directory names are not. status must be finalized for an episode to be trainable; success, the deliberate_failure tag and a _fail task name each zero the action-imitation loss.

Full schema — every field, unit, threshold, the time base and a runnable "read one episode" snippet — is docs/dataset_schema.md in the code repository.

Packaging

Loose — one directory per episode under 20260915_experiment{,_extra,_superseded}/, browsable in the file viewer and fetchable one episode at a time:

hf download armteam/hapticwam-rig-episodes --repo-type dataset --local-dir . \
    --include "20260915_experiment/ep_student_Carton_1789493413_000/*"

# one episode, resolved to the single archive that holds it, from a code checkout
python tools/hub/fetch_episode.py --dataset rig --episode ep_student_Carton_1789493413_000
python tools/hub/fetch_episode.py --dataset rig --episode first --samples   # the sample

Part of the HapticWAM release

Ten repos on the hub, gathered in the HapticWAM — ICRA 2027 collection.

Repo	Kind	Holds
armteam/hapticwam-teacher	model	the tactile-input teacher. Deployed checkpoint teacher_v6_simft/teacher_002000.pt; also holds the Cosmos prompt cache text_embeddings.pt
armteam/hapticwam-student	model	the distilled pad-free student, the model that runs on the rig. Deployed checkpoint hid_simft/student_001000.pt
armteam/hapticwam-baselines	model	the pi0.5, Diffusion Policy and X-VLA baselines at the deployed steps
armteam/hapticwam-ablations	model	every training arm that is not deployed, and the complete evaluation sweeps
armteam/hapticwam-teleop-dataset	dataset	the training corpus — 1,115 teleoperated episodes, packed per task
armteam/hapticwam-teleop-raw	dataset	the same teleoperation as loose, as-recorded sessions (provenance)
armteam/hapticwam-sim-episodes	dataset	Isaac Sim expert episodes, used for the sim fine-tune
armteam/hapticwam-rig-episodes ← you are here	dataset	the closed-loop rig takes the reported numbers are computed from
armteam/hapticwam-rollouts	dataset	policy-driven rollouts — the DAgger rounds and the deploy days
armteam/hapticwam-evidence	dataset	per-take evidence behind the paper's tables — scored CSVs, probe JSONs, figures

Code, training and deployment scripts: github.com/Advanced-Robotic-Manipulation/HapticWAM.

Licence

Data: CC-BY-4.0. The HapticWAM code and the model weights in the repos above: Apache-2.0.