map Open Source · Queued to study · 95/100

TUM-AVS/PlannerForge

The official implementation behind the PlannerForge paper — an agent pipeline that generates, retrieves, mutates, and scores autonomous-driving scenarios.

Python · Official research repository · EMNLP 2026Spatial & Simulation Systems
View on GitHub open_in_new

Activity

Linked directly from the accepted EMNLP 2026 paper.

What to Study

Scenario schema, tool adapters, retrieval/mutation strategy, and how planner benchmarking is scored.

Why It Matters

Strong bridge to simulation, robotics, GIS, and testing interests — official code that enables real domain adaptation.

Action

Inspect the scenario schema; map a logistics or facility-operations domain transfer.

Notes

  • Candidate for the Agentic Simulation Foundry experiment

AutoNateAI Experiments Using This

Proposed

Agentic Simulation Foundry

Does a generate → retrieve → mutate → execute → score pipeline transfer from autonomous driving to logistics, GIS, or facility-operations scenarios?

hub Spatial & Simulation Systems Read the Experiment arrow_forward
Proposed

Temporal GIS Narratives

Can sheaf/narrative structure preserve provenance and consistency in evolving geographic knowledge graphs better than event logs?

hub Spatial & Simulation Systems Read the Experiment arrow_forward

Part of

Forming

Spatial & Simulation Systems

Agentic scenario generation, GIS, and physical-world grounding — from autonomous driving benchmarks to logistics.

  • radarResearch Paper Radar
  • radarEvents & Build Radar
  • radarCalifornia Network Radar
Enter Project arrow_forward