- AutorIn
- Norman Seyffer
- Robert Schneider
- Titel
- Visibility-aware navigation for outdoor robots
- Untertitel
- Line-of-sight estimation using OSM and LiDAR
- Zitierfähige Url:
- https://nbn-resolving.org/urn:nbn:de:bsz:105-qucosa2-981301
- Konferenz
- Workshop on Autonomous Delivery and Service Robots on Pedestrian and Cycle Paths. Schkeuditz, Germany, 19.-20.05.2025
- Quellenangabe
- Collected Volume of the Workshop on Autonomous Delivery and Service Robots on Pedestrian and Cycle Paths
Erscheinungsort: Freiberg
Erscheinungsjahr: 2025 - Abstract (EN)
- Autonomous robots must not only perceive their surroundings reliably but also be visible to pedestrians and cyclists to ensure safe operations in urban environments. Static, semi-static and dynamic objects often block the line-of-sight between traffic participants. This paper presents a line-of-sight analysis that evaluates the visibility along a predefined 2 km test route in the city of Freiberg. The aim is to determine whether OpenStreetMap (OSM) data can provide sufficiently accurate information for predicting visibility conditions and assessing the need for sensor coverage. To assess whether freely available OpenStreetMap (OSM) data can be reliably used for field-of-view predictions, or if additional sensor-based mapping is necessary, OSM-based simulations are compared against real-world LiDAR point cloud data collected during a test drive. Based on the resulting visibility analysis, maximum safe velocities for the robot at each point along the route are proposed, ensuring that both traffic participants can come to a safe stop in the event of a sudden encounter. This enables adaptive speed planning depending on local visibility conditions, contributing to both operational safety and efficiency. Additionally, the findings can help identify areas where a robot should increase sensor coverage, or avoid certain locations due to poor visibility.
- Freie Schlagwörter (EN)
- visibility, viewablity, outdoor, autonomous robots, OSM, OpenStreetMap, LiDAR, sight, sightlines, lines-of-sight, raycasting, motion planning, point cloud, urban, speed planning, navigation, obstacle detection, pedestrian safety, human-robot interaction
- Klassifikation (DDC)
- 004
- Normschlagwörter (GND)
- Autonomer Roboter
- Navigation
- OpenStreetMap
- Lidar
- Unfallverhütung
- Herausgeber (Institution)
- Technische Universität Bergakademie Freiberg, Freiberg
- Version / Begutachtungsstatus
- publizierte Version / Verlagsversion
- URN Qucosa
- urn:nbn:de:bsz:105-qucosa2-981301
- Veröffentlichungsdatum Qucosa
- 24.09.2025
- Dokumenttyp
- Konferenzbeitrag
- Sprache des Dokumentes
- Englisch
- Lizenz / Rechtehinweis
CC BY 4.0