Occupancy Grids for Autonomous Driving

OccDrive – 1st Workshop on Occupancy Grids for Autonomous Driving

We are happy to announce that our workshop on Occupancy Grids for Autonomous Driving has been accepted by the 36th IEEE Intelligent Vehicles Symposium (IV). It will be part of a joint workshop on “Navigating Uncertainty: High-Integrity Localization, Mapping, and Perception for Autonomous Driving”.

Our workshop addresses all experts and students in robotics, autonomous systems, AI, and computer vision who are working on perception, mapping, and sensor fusion.

Join and discuss with us on June 22, 2025 in Cluj-Napoca, Romania.
Register here for the 26th IEEE IV Symposium

About the Workshop

Autonomous systems, including robots and self-driving vehicles, rely on diverse perception algorithms to understand and interact with their environments. Among these, occupancy grids are widely used for spatial representation.

Traditionally, occupancy grids discretize the environment into a 2D or 3D grid, where each cell carries a probability indicating whether it is occupied, free or unknown. Furthermore, it can be extended with semantic labels or dynamic states. This fundamental representation is essential for downstream tasks such as prediction and planning for autonomous driving.

Occupancy grids, though well-established, have evolved significantly with advancements like dynamic occupancy grids and end-to-end 3D occupancy prediction. Occupancy grids are also widely used in sensor fusion frameworks, both in traditional architectures and in combination with deep learning approaches. All these sub-fields provide exciting areas for research.

Additionally, the emergence of foundation models and the shift toward fully differentiable autonomous driving frameworks have introduced further research directions, influencing how occupancy-grid based perception evolves.

This workshop seeks to unite researchers to exchange ideas and share recent developments and results in this field. It also aims to discuss the future of occupancy grid mapping—exploring, for example, whether occupancy grids will remain standalone algorithms or increasingly integrate with broader aspects of autonomous systems.

June 22, 2025
Cluj-Napoca, Romania
Grand Hotel Italia, Bernini Room 80

Workshop Topics

  • 2D/3D (Dynamic) Occupancy Grids
  • End-to-End Occupancy Grid Prediction
  • Foundation Models for Occupancy Grids
  • Occupancy-grid based Prediction and Planning
  • Efficient computation of Occupancy Grids
  • Datasets for Occupancy Grids
  • LLMs/VLMs for Occupancy Grid Mapping
  • Birds-eye-view based perception/prediction/planning
  • Multi-modal sensor fusion for Occupancy Grids

Speakers

murray
Shane Murray
NVIDIA
dorn
Dr. Gerhard Dorn
Virtual Vehicle Research Center
valada_page
Prof. Abhinav Valada
University of Freiburg

Join and discuss with us on June 22, 2025.

Organization TEAM

The 1st Workshop on Occupancy Grids for Autonomous Driving is jointly organized by:

ronecker
Max Peter Ronecker
SETLabs Research GmbH / Graz University of Technology
neumann
Felix Neumann
Siemens AG
diaz_2
Xavier Diaz
SETLabs Research GmbH
watzenig-8
Prof. Daniel Watzenig
Graz University of Technology
burschka
Prof. Darius Burschka
Technical University of Munich