Description
The robot Unimog U5023 is a Mercedes Benz Special Trucks vehicle with high-level terrain mobility and many possible utilization areas. It is characterized by extreme off-road capability and features like differential locks, a central tire inflation system, a high number of gears, frame torsion, centerline torsion, and portal axle. Therefore, the applied research benefits from its ability to access rough terrain not included in the previous research programs. Accordingly, a research focus is on developing robust and secure commercial vehicles' navigation solutions in rough and unstructured terrain. Compared to on-road applications, off-road navigation is still an unsolved research problem with high technical barriers. Nevertheless, the legal regulations for autonomous commercial vehicles in unstructured terrain are typically less strict, and concepts commercialize more easily.
The Unimog's central research topics are the progress of robust perception systems, novel navigation methodologies in challenging environments, and the deployment of autonomous tooling in harsh or at least complex environments. Consequently, the vehicle's perception provides, for instance, detailed information about object detections, classifications, scene context, the evaluation of passages, plus obstacles. On top of that, there are biologically motivated methods for localization and general quality assessment of the sensor system. Especially in unstructured environments, a robot system is often influenced and disturbed by noisy and faulty sensor measurements.
Therefore, project partners and lab members extended a standard Unimog 5023 for autonomous deployment, including a new steering and braking system, computing capabilities, and sensor systems. Furthermore, a complex simulation of the Unimog was built, including all the vehicle's extras and characteristics. On account of this, it is possible to simulate risky maneuvers in heavy terrain before testing them in real life.
Videos
Publications
- Evaluating the Robustness of Off-Road Autonomous Driving Segmentation against Adversarial Attacks. A Dataset-Centric analysis.
arXiv, (2024) - Evaluation of Robustness of Off-Road Autonomous Driving Segmentation against Adversarial Attacks. A Dataset-Centric Study.
Proceedings of the 23rd International Conference on Autonomous Agents and Multiagent Systems, Nr. 3, S. 22372239. (2024) - Exploring Image Fusion Techniques for Off-Road Semantic Segmentation in Harsh Lighting Conditions. A Multispectral Imagery Analysis.
Proceedings of the 21st International Conference on Ubiquotous Robots (UR2024), (2024) - Implementation of Off-Road Panoptic-Segmentation under dynamic lighting conditions.
Proceedings of 23rd International Conference on Machine Learning and Applications (ICMLA 2024) (To appear), (2024) - Understanding nature step-by-step. towards off-road environment perception for autonomous driving.
Proceedings of 6th Young Researchers Symposium (YRS 2024), (2024) - Aktuelle Forschung im Bereich des autonomen Fahr- und Arbeitsbetriebs von Nutzfahrzeugen.
Jahresmagazin AUTOMATION + ROBOTIK, S. 48 - 53. (2023) - Global map generation using local feature grid maps for autonomous vehicles in frequently changing off-road environments.
18th International Conference on Intelligent Autonomous Systems (IAS 2023), (2023) - Navigating Off-roads. Using Deep neural Networks for Rock detection in off-road Autonomous Driving with Unimog.
20th International Conference on Ubiquitous Robots (UR 2023), (2023) - Off-Road Trajectory Generation.
(2023) - Reverse-Engineering of Behavior-Based Robot Components.
18th International Conference on Intelligent Autonomous Systems (IAS 2023), (2023) - Robust Autonomous Off-Road Navigation Incorporating Self-Learned, Higher-Level Map Knowledge.
(2023) - Simulation-based tool for testing autonomous vehicles.
(2023) - Simultaneous top-down and bottom-up perception for autonomous off-road driving.
2023 International Conference on Robotics and Automation (ICRA), (2023) - Cognitive Processing in Behavior-Based Perception of Autonomous Off-Road Vehicles.
(2022) - Disturbance Detection in LiDAR Data.
(2022) - Disturbance and Particle Detection in LiDAR Data.
48th Annual Conference of the IEEE Industrial Electronics Society (IECON 2022), (2022) - Filtering of Laser Scanner Point Clouds Tainted by Rain or Snow.
(2022) - Human-inspired cognitive processing for robust autonomous off-road driving.
5th Young Reseachers Symposium 2022, S. 35. (2022) - Traction Control of Autonomous Off-road Vehicle using Semantic Segmentation and Offline-maps.
(2022) - Traction optimization for robust navigation in unstructured environments using deep neural networks on the example of the off-road truck Unimog.
17th International Conference on Intelligent Autonomous Systems – IAS-17, S. 191 - 208. (2022) - Tree-Based Localization and Mapping in Rough Forest Environments for Autonomous Off-Road Vehicles.
(2022) - Universelle Nutzfahrzeugkonzepte zum autonomen Arbeiten in herausfordernden Umgebungen. Ein Zwischenbericht zum Projekt »Autonome Arbeitsmaschine für den Einsatz in Anwendungen wie Katastrophenschutz/ Feuerwehr, Forstwirtschaft und Weinbau«..
CVC News, Vol. 2, S. 20 - 24. (2022) - Autonome Navigation und Umfelderkennung eines Unimogs in unstrukturiertem Terrain.
Tagungsband DWT-SGW Forum Unmanned Systems VIII, S. 30. (2021) - Autonomes Arbeiten mit dem Unimog. Umrüstung für Arbeitseinsätze im Katastrophenschutz, bei Feuerwehr, Forstwirtschaft, Weinbau und Deponie.
CVC News, Vol. 2, S. 8 - 10. (2021) - Data-fusion for robust off-road perception considering data quality of uncertain sensors.
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), S. 6853 - 6860. (2021) - Towards effective Rock detection in Off-Road environments for Autonomous Navigation.
(2021) - Advanced scene aware navigation for the heavy duty off-road vehicle Unimog.
The 9th International Conference on Advanced Concepts in Mechanical Engineering – ACME 2020. IOP Conference Series, Vol. 997, S. 1 - 18. (2020)
DOI: 10.1088/1757-899x/997/1/012093 - Autonome Off-Road Navigation von Nutzfahrzeugen. Erfahrungsbericht und Testergebnisse des CVC-Leitprojekts zum autonomen Fahrbetrieb von Nutzfahrzeugen im Off-Road-Bereich.
CVC News, Vol. 1, S. 4 - 10. (2020) - Autonomous Off-Road Navigation using Near-Feature-Based World Knowledge Incorporation on the Example of Forest Path Detection.
Preprint submitted to Robotics and Autonomous Systems, (2020) - Wenn Autos fahren wie von Geisterhand. Off-Road-Bereich wird zur Herausforderung / Nutzen wäre aber enorm.
KOMMUNALtopinform, Vol. 4, S. 62 - 63. (2020) - Behavior-Based Control for Safe and Robust Navigation of an Unimog in Off-Road Environments.
Commercial Vehicle Technology 2018. Proceedings of the 5th Commercial Vehicle Technology Symposium – CVT 2018, S. 63 - 76. (2018) - Behavior-Based Low-Level Control for (semi-) Autonomous Vehicles in Rough Terrain.
Proceedings of ISR 2018, S. 386 - 393. (2018) - Local Behavior-Based Navigation in Rough Off-Road Scenarios based on Vehicle Kinematics.
2018 IEEE International Conference on Robotics and Automation (ICRA), S. 719 - 724. (2018)
DOI: 10.1109/ICRA.2018.8460631 - Modelling and Simulation of Behaviour-Based Differential and Slippage Control for Unimog.
(2018) - Multi Feature Maps for Autonomous Off-Road Navigation in Rough Environments.
(2018) - Ontologies for Situation-Aware, Autonomous Navigation in Challenging Off-Road Environments.
(2018) - Autonomie in unwegsamem Gelände. Aktuelle Zwischenergebnisse des CVC-Leitprojekts zum autonomen Fahrbetrieb von Nutzfahrzeugen im Off-Road Bereich.
CVC News, Vol. 2, S. 8 - 11. (2017) - Behavior-Based Gear Control for Rough Off-Road Environments.
(2017) - Behavior-Based Navigation for Stuck Vehicles in Rough Off-Road Environments.
(2017) - Geländerekonstruktion basierend auf 3D-Punktwolken.
(2017)