nuPlan is the world's first large-scale planning benchmark for autonomous driving.

nuPlan is the world's first large-scale planning benchmark for autonomous driving. While there is a growing body of ML-based motion planners, the lack of established datasets, simulation frameworks and metrics has limited the progress in this area. Existing benchmarks (Argoverse, Lyft, Waymo) for autonomous vehicle motion prediction have focused on short-term motion forecasting of other agents, rather than long-term planning of the ego vehicle. This has led previous works to use open-loop evaluation with L2-based metrics, which are not suitable for fairly evaluating long-term planning. Our benchmark overcomes these Limitation by providing a training framework to develop machine learning based planners, a lightweight closed-loop simulator, motion-planning specific metrics and an interactive tool to visualize the results.

NuPlan mini sample (VADv2-ready)

This dataset contains a single NuPlan mini log converted into the VADv2 training/evaluation format.

  Contents

nuplan_data/ maps/ # Full NuPlan map bundle (gpkg, npy, json) nuplan-v1.1/ sensor_blobs/2021.05.12.22.00.38_veh-35_01008_01518/ CAM_* folders + MergedPointCloud/ # Cameras and fused lidar frames splits/ mini/2021.05.12.22.00.38_veh-35_01008_01518.db # Mini split SQLite DB… See the full description on the dataset page: https://huggingface.co/datasets/runfuture/nuplan-mini-sample.

Dataset

This dataset was created by sisi4376

Contents

We provide a standardized graph dataset for traffic based on the large-scale NuPlan v1.1 dataset, converted to a PyTorch-Geometric dataset using our CommonRoad-Geometric tool. The dataset is collected from 3 megacities: Singapore, Boston and Pittsburgh.

Scene Description NuPlan Dataset

A dataset of natural language descriptions generated from autonomous driving simulator states, paired with visualization images of driving scenarios.

  Overview

The scene_description_nuplan_dataset contains natural language descriptions of driving scenarios generated from the V-Max simulator using the NuPlan dataset. This dataset provides paired data of:

Natural language descriptions of driving scenes Corresponding visualization images at… See the full description on the dataset page: https://huggingface.co/datasets/ac4462/SceneDescriptionNuPlan.

ChiChi2000/Nuplan-QA dataset hosted on Hugging Face and contributed by the HF Datasets community

We curated Nexus_data, which consists of complex corner cases based on NuPlan dataset. It covers 540 hours of simulated data, including high-risk interactions such as cut-in, sudden braking, and collision. More details can be found in Decoupled Diffusion Sparks Adaptive Scene Generation, accepted by ICCV,2025.

Dataset

This dataset was created by Kanishkha Jaisankar

Released under MIT

Contents