UGV
FoMo
Multi-season robot localization and SLAM
A boreal-forest UGV dataset with over 64 km of repeated seasonal routes, paired lidars, stereo and rear camera imagery, radar, inertial measurements, audio, power and weather metadata, and GNSS-derived position references.
Access links
Primary places to inspect, download, read about, or reproduce the dataset.
Quick facts
- Vehicle type
- UGV
- Environment
- Outdoor
- Data origin
- Real-world
- Sensor count
- 12
- Ground truth
- Available
- Calibration
- Full Raw Calibration Data
- Annotations
- Not reported
Sensor overview
High-level sensor availability before the detailed sensor records below.
Ground truth
Reference scope, method, rate, coverage, and provenance.
| Scope | Method | Reference system | Count | Rate | Coverage |
Reference provenance
|
|---|---|---|---|---|---|---|
| 3D position | Other | Three-rover PPK GNSS with static base and CORS correction · Emlid M2 | 3 onboard GNSS reference receivers | Not reported | Full |
Reference provenance
Hybrid
|
|
||||||
Calibration and synchronization
Reported calibration level, reproducibility signals, and supporting notes.
- Level
- Full Raw Calibration Data
- Processed parameters
- Available
- Raw calibration data
- Available
- Calibration targets
- Checkerboard, retro-reflective circular target and AprilTag validation screen
Reproducible calibration data
Raw calibration data is reported, so users can reproduce or inspect the calibration process.
Notes
Per-recording calibration parameters and downloadable calibration sequences. Allan variance IMU characterization uses 24-hour data. Basler intrinsics use a checkerboard; ZED X intrinsics are manufacturer-provided. A Hesai PandarXT-32 is used to bridge Basler extrinsics and is calibration equipment, not a normal route lidar. Radar alignment optimizes x, y and yaw.
Known or intentional limitations
Reported constraints and characteristics to check before using the dataset.
Ground truth contains position only; identity quaternions do not encode measured attitude.
Reference covariances underestimate true position errors, particularly under forest canopy.
Some route-deployment pairs are absent and the winter Green route changes.
Metadata CSVs are excluded from generated MCAP bags.
Basler auto-exposure failed in one cold recording.
Sensor scans are not compensated for ego motion.
Cameras
zed_x
- Model
- ZED X
- Setup
- Stereo
- Count
- 1
- Effective cameras
- 2
- Modality
- Not Reported
- Resolution
- 1920 x 1200
- Rate
- 10 Hz
- Shutter
- Not Reported
- Lens type
- Not Reported
- HFOV
- Not reported
- VFOV
- Not reported
- Notes
- Rectified PNG images. Driver timestamps have an estimated 51.48 ms mean delay with 8.78 ms standard deviation. The 56 Hz setting belongs to calibration, rather than normal route recordings.
basler
- Model
- Basler ace2
- Setup
- Mono
- Count
- 1
- Effective cameras
- 1
- Modality
- Not Reported
- Resolution
- 1920 x 1200
- Rate
- 10 Hz
- Shutter
- Not Reported
- Lens type
- Not Reported
- HFOV
- Not reported
- VFOV
- Not reported
- Notes
- Rear-facing wide-lens images are undistorted PNGs. Intrinsic calibration emphasizes the ground behind the robot. One cold-weather recording has an auto-exposure failure producing a white image.
IMUs
Xsens MTi-30
- Model
- Xsens MTi-30
- Count
- 1
- Accel / gyro
- Not reported
VectorNav VN100
- Model
- VectorNav VN100
- Count
- 1
- Accel / gyro
- Not reported
GNSS
Not reported
LiDAR
RoboSense Ruby Plus
- Model or name
- RoboSense Ruby Plus
- Count
- 1
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- Not reported
- Range
- Not reported
Leishen LS128S1
- Model or name
- Leishen LS128S1
- Count
- 1
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- Not reported
- Range
- Not reported
Radar
Navtech CIR-304H
- Count
- 1
- Released data
- Scanning power images
- Frame rate
- 4 Hz
- Doppler velocity
- Not reported
- Elevation angle
- Not reported
- Configured range
- Not reported
- Field of view
- Not reported
- File format
- Oxford-format polar PNG
- Notes
- Two-dimensional polar scans with per-azimuth timestamps and rotational encoder values. No ego-motion compensation. SDK conversion to Cartesian images does not establish released point clouds or Doppler measurements.
Pitot tube
Not reported
Additional sensors
Barometer
- Model
- Adafruit DPS310
- Count
- 1
- Rate
- 100 Hz
- Notes
- Pressure data are in metadata CSVs, which are excluded from generated MCAP files.
Other sensor
- Name
- Left and right microphones
- Count
- 2
- Rate
- 16000 Hz
- Notes
- Independent left and right WAV channels, stored as one-second snippets; Table 2 and Figure 6 disagree on model.
Odometry
- Name
- Differential-drive wheel odometry
- Count
- 1
- Notes
- Odometry computed from the ideal differential-drive model.
Wheel encoder
- Model
- Hall effect sensors
- Count
- 2
- Rate
- 4 Hz
Annotations
Reported annotation availability and task support.
- Availability
- Not reported
- Format
- Not reported
- Class count
- Not reported
Semantic segmentation
Not reported
Instance segmentation
Not reported
Object detection
Not reported
Optical flow
Not reported
Depth ground truth
Not reported
Citation
Citation key and BibTeX kept at the end of the page for reference.
Not reported
Dataset corrections
Notice a specific issue after reviewing the page?
Report an inaccuracy