From September 27 to October 1, 2026, the TUM Autonomous Vehicle Systems Lab will be represented at the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026) in Pittsburgh, USA.

With five publications, two co-organized workshops, one invited talk, and several workshop contributions, our team will present current research in the areas of planning, control, scenario generation, perception, and autonomous racing.

Five Publications from the TUM AVS Lab

At IROS 2026, our team will present the following works:

1. Differentiable Weights-Varying Nonlinear MPC Via Gradient-Based Policy Learning: An Autonomous Vehicle Guidance Example
   Felix Jahncke, Baha Zarrouki, Mattia Piccinini, Jovin D'sa, David Isele, Sangjae Bae, and Johannes Betz
   📄 IEEE Robotics and Automation Letters (RA-L)
   📅 September 30, 3:16 PM · 📍 Room 320
2. G²DP: Diffusion Planning with Spatio-Temporal Grid Guidance
   Hang Yu, Marc Kaufeld, and Johannes Betz
   📅 September 28, 9:46 AM · 📍 Room 321
3. GaussianFusionOcc: A Seamless Sensor Fusion Approach for 3D Occupancy Prediction Using 3D Gaussians
   Tomislav Pavković and Johannes Betz
   📅 September 28, 2:56 PM · 📍 Rooms 409/410
4. Chat2Scenic: An Iterative RAG-Based Framework for Scenario Generation in Autonomous Driving
   Yuan Gao, Mattia Piccinini, and Johannes Betz
   📅 September 29, 3:25 PM · 📍 Rooms 401/402
5. A²RL V_max: The A²RL Autonomous Racing Dataset for Long-Range, High-Speed Perception and Multi-Vehicle Interaction
   Marvin Klemp and Johannes Betz
   📅 September 30, 10:27 AM · 📍 Rooms 317/318

Co-Organized Workshops

Our researchers are also involved in organizing two workshops.

The 16th Workshop on Planning, Perception and Navigation for Intelligent Vehicles (PPNIV 2026) is organized by Junmin Wang, Johannes Betz, Yufeng Yue, and Hyun Myung. The workshop focuses on current research in planning, perception, and navigation for intelligent vehicles and brings together international experts from robotics and autonomous driving.
📅 September 27, 1:30–5:30 PM
📍 Room 412

The workshop “AI Meets Control for Resilient and Reliable Robot Planning (AIM-Ctrl)” is organized by Felix Jahncke, Yuan Gao, Mattia Piccinini, Johannes Betz, Daniele Meli, Francesco Trotti, and Alessandro Farinelli. The workshop focuses on combining artificial intelligence methods with model-based control and planning for robust, resilient, and reliable robotic systems.
📅 October 1, 8:30 AM–12:30 PM
📍 Rooms 408/409

Invited Talk

As part of PPNIV 2026, Mattia Piccinini will give the invited talk “Differentiable Physics-Informed Learning and Control in Autonomous Vehicle Racing.” The talk will provide insights into the integration of physics-based models, differentiable optimization methods, and learning-based approaches for autonomous race vehicle control, including our work on Diff-WMPC.
📅 September 27, 3:30–3:50 PM
📍 Room 412

Workshop Contributions

Several of our works will also be presented as part of workshops:

• Diff-WMPC — Felix Jahncke at the Hybrid Architectures for Embodied Autonomy (BLPC) workshop
  📅 September 27, 1:30–5:30 PM · 📍 Rooms 303/304
• Chat2Scenic — Yuan Gao at the AI Meets Autonomy workshop
  📅 October 1, 8:30 AM–12:30 PM · 📍 Rooms 319/320
• JAMP — Marc Kaufeld at the AIM-Ctrl workshop
  📅 October 1, 8:30 AM–12:30 PM · 📍 Rooms 408/409

We look forward to presenting our latest research at IROS 2026, engaging with the international robotics community, and meeting colleagues and collaborators in Pittsburgh.