This research data repository was developed and curated by Prof. Dr.-Ing. Patrick Wolf and Prof. Dr. Karsten Berns and is supported by the Rhineland-Palatinate Ministry of Economic Affairs, Transport, Agriculture and Viticulture (MWVLW RLP).

The repository provides multimodal data records that were collected as part of extensive off-road field tests with a Unimog vehicle under demanding real-life operating conditions. The records cover a variety of operating scenarios, including steep unpaved climbs, gravel terrain, forest trails, industrial areas and complex off-road intersections under varying environmental and mechanical conditions.

The data sets focus on robust terrain perception and vehicle mobility analysis. Obstacle detection, dynamic elevation changes, strong vehicle vibrations and challenging lighting changes due to dense vegetation and uneven terrain surfaces are recorded. The experimental recordings include uphill and downhill drives with gradients between 15° and 60° and enable a detailed investigation of vehicle behavior and sensor performance under extreme off-road conditions.

The synchronized sensor system includes:

• 3D LiDAR laser scanner

• camera systems

• IMU (Inertial Measurement Unit)

• GNSS systems

The data sets provided are used for research in the fields of autonomous off-road navigation, sensor fusion, environmental perception, localization and terrain-adaptive mobility systems.

This recording documents the vehicle traveling across rough, unpaved gravel terrain along an upward trajectory toward a T-intersection. The surroundings include characteristic roadside features, such as scattered rock fragments and vegetation of varying density. Of particular note is the vehicle’s approach to a mound of earth, which is detected and represented as a primary obstacle within the 3D laser point cloud data.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0001

DOI:https://doi.org/10.26204/data/15

Plain text: Wolf, Patrick and Berns, Karsten, "Uphill Gravel Terrain and Obstacle Detection," 2026, doi:10.26204/DATA/15.

BibTeX: t3://file?uid=15147

Scenario Description

The vehicle navigates along a downward-sloping path through a Y-intersection and follows the left off-road route. The scenario documents the transition from rough terrain to an industrial area of a waste disposal center, with structural environmental data being captured despite significant camera-based image distortions caused by vehicle vibrations. The recording ends upon reaching level terrain, as the vehicle passes construction vehicles and various technical facilities on its way back to the starting point.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0002

DOI:https://doi.org/10.26204/data/16

Plain text: Wolf, Patrick and Berns, Karsten, "Downhill Off-Road Descent and Industrial Structural Observation," 2026, doi:10.26204/DATA/16

BibTeX: t3://file?uid=15137

Scenario Description

This recording documents the vehicle’s operational capability under extreme off-road conditions and focuses specifically on several uphill maneuvers on slopes ranging from 15 to 60 degrees. The scenario includes two separate driving cycles over steep, unpaved inclines to capture topographic data and sensor responses under significant mechanical stress. The environment is characterized by loose ground and drastic changes in elevation, providing a high-stress test case for 3D laser point cloud obstacle and terrain mapping.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0003

DOI:https://doi.org/10.26204/data/17

Plain text: Wolf, Patrick and Berns, Karsten, "Extreme Hill Climbing: High-Angle Incline Performance," 2026, doi:10.26204/DATA/17

BibTeX:t3://file?uid=15138

Scenario Description

This recording documents the vehicle navigating a narrow path with dense vegetation on an uphill trajectory toward a designated campsite. The scenario highlights off-road driving conditions through forested areas that require precise spatial awareness. The 3D laser data illustrates the proximity of the surroundings and highlights the sensors’ ability to map narrow paths during steep ascents.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0004

DOI: https://doi.org/10.26204/data/18

Plain text: Wolf, Patrick and Berns, Karsten, "Extreme Uphill to Tent: Narrow Path Navigation," 2026, doi:10.26204/DATA/18

BibTeX:t3://file?uid=15139

Scenario Description

This recording documents the vehicle navigating a forest trail characterized by uneven surfaces, varying textures, and slight elevation changes. The dense canopy poses significant optical challenges, exposing the visual sensors to abrupt, high-contrast transitions between bright sunlight and deep shadows. The scenario captures sensor performance under these dynamic lighting conditions as the vehicle traverses the rugged terrain toward a structural landmark.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0005

DOI:https://doi.org/10.26204/data/19

Plain text: Wolf, Patrick and Berns, Karsten, "Forest Path to Windmill: Variable Lighting and Terrain," 2026, doi:10.26204/DATA/19

BibTeX:t3://file?uid=15143

Scenario Description

This recording captures the vehicle navigating steep descents along an off-road circuit. The driving scenario maintains a constant speed profile while traversing rough, unpaved terrain characterized by loose gravel and rocky surfaces. Located adjacent to a construction site, the dataset provides sensor observations of the topographic gradient and the surrounding structural environment during controlled, extreme downhill maneuvers.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0006

DOI: https://doi.org/10.26204/data/20

Plain text: Wolf, Patrick and Berns, Karsten, "Round Track Extreme Downhill: Constant Speed Descent," 2026, doi:10.26204/DATA/20

BibTeX:t3://file?uid=15145

Scenario Description

This recording captures the vehicle as it traverses steep, artificial earth mounds and documents both uphill and downhill drives. The operating environment consists of extremely rough, loose earth terrain that generates significant amounts of airborne particles. The resulting dust actively covers and obscures the optical sensors, providing a specialized dataset for evaluating camera performance and the impairment of perception under poor visibility conditions.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0007

DOI: https://doi.org/10.26204/data/21

Plain text: Wolf, Patrick and Berns, Karsten, "Manmade Terrain Navigation and Visual Obscuration," 2026, doi:10.26204/DATA/21

BibTeX:t3://file?uid=15141

Scenario Description

This recording documents vehicle operation in muddy terrain interspersed with debris, featuring significant pools of water and uneven topographical challenges. The driving scenario is conducted under foggy weather conditions, resulting in slightly impaired visual visibility. The dataset captures a downhill step maneuver followed by a drive over coarse scree, culminating in a failure to climb an uphill slope, thereby providing critical sensor telemetry for low-friction, complex off-road environments.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0008

DOI:https://doi.org/10.26204/data/22

Plain text: Wolf, Patrick and Berns, Karsten, "Muddy Rubble Transit and Uphill Failure," 2026, doi:10.26204/DATA/22

BibTeX:t3://file?uid=15146

Scenario Description

This recording documents the vehicle navigating a moderate circuit characterized by rough, muddy terrain covered with rocks and debris. The driving scenario includes a descent followed by an attempted ascent on loose, yielding ground. The environment provides a test site with low traction, allowing the onboard sensors—including the 3D laser—to map the complex topography and the vehicle’s spatial dynamics during the challenging climbing maneuver.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0009

DOI:https://doi.org/10.26204/data/23

Plain text: Wolf, Patrick and Berns, Karsten, "Moderate Round Track: Muddy Terrain and Hill Climb," 2026, doi:10.26204/DATA/23

Citation:t3://file?uid=15144

Scenario Description

The recording captures the vehicle’s operation along a muddy, uneven circular track located in an open landfill or construction site with no adjacent structures or vegetation. The primary driving maneuver involves climbing an artificial hill under poor lighting conditions. The dataset illustrates significant impairment of the optical sensors caused by suboptimal lighting conditions and physical obstruction by water or dust on the camera lens.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0010

DOI:https://doi.org/10.26204/data/24

Plain text: Wolf, Patrick and Berns, Karsten, "Round Track: Muddy Terrain and Optical Degradation," 2026, doi:10.26204/DATA/24

BibTeX:t3://file?uid=15148

Scenario Description

This recording documents a scenario involving a stationary vehicle in which the entire sensor suite, including all active cameras, is utilized. The dataset captures dynamic interactions in the immediate vicinity and specifically records pedestrians moving around the stationary ego vehicle. In addition, a second vehicle is tracked as it repeatedly crosses the area near a structural reference point and passes the ego vehicle at specific intervals to provide localized data for motion and obstacle detection.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0011

DOI:https://doi.org/10.26204/data/25

Plain text: Wolf, Patrick and Berns, Karsten, "Full Sensor Suite: Pedestrian and Vehicle Interaction," 2026, doi:10.26204/DATA/25

BibTeX:t3://file?uid=15142

Scenario Description

This scenario captures the vehicle traversing a challenging off-road environment consisting of loose gravel, significant amounts of boulders, and sparse vegetation. Data acquisition utilized the complete sensor suite, which includes 3D lasers, safety lasers, and stereo cameras. The driving maneuvers include navigating over substantial rock piles on both uphill and downhill slopes and provide a comprehensive multimodal dataset for perception in rugged terrain.

Quick-access links:

https://fdm-fallback.uni-kl.de/RPTU/FB/Computer Science/AG-Wolf/0012

DOI: https://doi.org/10.26204/data/26

Plain text: Wolf, Patrick and Berns, Karsten, "Full Sensor Suite in Rough Terrain," 2026, doi:10.26204/DATA/26

BibTeX:t3://file?uid=15140