THULab/aloha_static_battery

ALOHA Static Battery (TsFile)

Apache TsFile version of lerobot/aloha_static_battery.

The source dataset was created with LeRobot and is licensed under Apache 2.0. It contains ALOHA robot demonstrations for the task "Place the battery into the slot of the remote controller."

Modalities: Time-series. The original camera streams remain available in the source Hugging Face dataset; this repository stores the numeric robot state, action, frame metadata, task index, and episode tags as Apache TsFile.

Source Dataset

• Original dataset: lerobot/aloha_static_battery
• Codebase version: LeRobot v3.0
• Robot type: unknown
• Split: train (0:49)
• Episodes: 49
• Frames: 29,400
• Sampling rate: 50 fps
• Tasks: 1
• Source data path: data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet
• Source video path: videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4

Converted Files

• TsFile: data/aloha_static_battery.tsfile
• Rows: 29,400
• Table name: aloha_static_battery
• Time precision: milliseconds
• Mirrored metadata: meta/, with meta/info.json updated for the TsFile artifact

Schema

Time is generated as round(timestamp * 1000) milliseconds. Time restarts within each episode, and devices are identified by the original LeRobot tag columns.

• TAG columns: episode_index, task_index
• FIELD metadata columns: frame_index, sample_index, next_done
• FIELD vectors: observation_state_0..observation_state_13 and action_0..action_13

next_done is the episode-termination flag. The converted data has 49 true flags, one terminal frame per episode.

Conversion Notes

• The source timestamp column is dropped after being mapped to Time; it is recoverable as Time / 1000 seconds.
• The source index column is renamed to sample_index.
• The source next.done column is renamed to next_done.
• Vector columns are flattened by preserving the full source column name, replacing . with _, and appending the element index.
• Source video features are not uploaded here: observation.images.cam_high, observation.images.cam_left_wrist, observation.images.cam_low, and observation.images.cam_right_wrist. Use the original dataset's videos/ tree for the visual streams.
• Aside from the redundant timestamp column and omitted video files, no numeric time-series rows are intentionally dropped.

Read Example

from tsfile import TsFileReader

path = "data/aloha_static_battery.tsfile"

with TsFileReader(path) as reader:
    tables = reader.get_all_table_schemas()
    print(tables.keys())