Download README.md from BitRobot/FrodoBots-Mini-4K: direct link, hf CLI and curl.
- Browser
- Download file 4.22 kB
-
https://huggingface.co/datasets/BitRobot/FrodoBots-Mini-4K/resolve/df0dc9a82082fa917337405686594de2f7b07564/README.md
- Command line
-
hf download hf://datasets/BitRobot/FrodoBots-Mini-4K@df0dc9a82082fa917337405686594de2f7b07564/README.md
-
curl -L -o README.md https://huggingface.co/datasets/BitRobot/FrodoBots-Mini-4K/resolve/df0dc9a82082fa917337405686594de2f7b07564/README.md
license: cc-by-sa-4.0
pretty_name: FrodoBots-Mini-4K
task_categories:
- robotics
- other
tags:
- robotics
- teleoperation
- autonomous-driving
- sidewalk-robot
- multimodal
- video
- gps
- imu
size_categories:
- n>10K
FrodoBots-Mini-4K
~4,000 hours of real-world teleoperation data from Earth Rover Mini / Mini+ sidewalk robots, collected across 30+ countries. Each ride bundles synchronized camera video (front, and rear when available), two audio channels, and time-aligned GPS, IMU, and drive/control (DRV) streams.
This is the third public FrodoBots dataset, following
BitRobot/FrodoBots-2K and
BitRobot/Berkeley-FrodoBots-7K.
Like the 2K release it ships raw, unannotated per-ride folders (no task labels) — packaged here as
tar shards hosted directly on the Hub.
At a glance
| Rides | 14,141 |
| Robots | 453 |
| Countries | 30+ |
| Date range | 2024-12 → 2026-07 |
| Hours (teleoperation session time) | 3,965 h (same metric as FrodoBots-2K/7K) |
| Front-camera video | 3,231 h |
| Rear-camera video | 1,909 h (73.9% of rides have a rear camera) |
| Total camera video | 5,140 h |
| Size | ~2.0 TB |
Every ride includes DRV + GPS + IMU + at least the front (or rear) camera with generated per-frame timestamps.
Layout
data/
minirover-0000.tar # tar shards, <= ~1.9 GB each, each holds many complete ride folders
minirover-0001.tar
...
metadata.parquet # one row per ride (see below)
Each ride folder inside a shard:
ride_<rideId>_<botRef>_<UTCstart>/
control_data_<id>.csv # DRV: linear, angular, rpm_1..4, timestamp (epoch SECONDS)
gps_data_<id>.csv # latitude, longitude, timestamp (epoch MS); (1000,1000)=no fix
imu_data_<id>.csv # compass, accelerometer, gyroscope (JSON arrays), timestamp (MS)
front_camera_timestamps_<id>.csv # frame_id, timestamp (epoch SECONDS)
rear_camera_timestamps_<id>.csv # present when a rear stream exists
mic_audio_timestamps_<id>.csv # robot microphone
speaker_audio_timestamps_<id>.csv # operator speaker
recordings/ # HLS: *.ts segments + *.m3u8 playlists (front/rear video, audio)
metadata.parquet columns: folder, ride_id, device_ref_id, start_utc, source, db_dur_sec, n_epochs, hardware_version, drive_mode, country, city, front_footage_sec, rear_footage_sec, has_rear_camera, has_control, has_gps, has_imu, has_front_ts, has_rear_ts, has_mic_ts, has_speaker_ts, shard.
Usage
from huggingface_hub import hf_hub_download
import pandas as pd, tarfile
meta = pd.read_parquet(hf_hub_download("BitRobot/FrodoBots-Mini-4K", "metadata.parquet", repo_type="dataset"))
row = meta.iloc[0]
shard = hf_hub_download("BitRobot/FrodoBots-Mini-4K", f"data/{row.shard}", repo_type="dataset")
with tarfile.open(shard) as t:
t.extractall("out", members=[m for m in t.getmembers() if m.name.startswith(row.folder)])
# -> out/<ride folder>/ with CSVs + recordings/
See earth_rover_mini_dataset.ipynb for stats, distributions, and a full "load one ride" example
(aligning a camera frame to its GPS/IMU/DRV readings).
Notes
- Hours: the headline counts teleoperation session duration (session end − start), the same basis as FrodoBots-2K/7K. Actual recorded video is 3,231 h front / 5,140 h total (front+rear); a small number of sessions were left open and slightly inflate raw session time.
- Recordings are de-vendored (Agora tags stripped) HLS.
.m3u8playlists carry the wall-clock start (TIME=) and per-segment durations; per-frame timestamps are provided in the*_timestamps_*.csvfiles. - GPS retains
(1000,1000)rows meaning "no fix" — filter these out before use. countryis from the robot registry; robots without an entry areUnknown.- Audio-only sessions (camera never streamed) are excluded.
License
Released under CC BY-SA 4.0, matching FrodoBots-2K and Berkeley-FrodoBots-7K. Please cite FrodoBots when using this dataset.