Towards learning-based planning:The nuPlan benchmark for real-world autonomous driving

TL;DR

The paper introduces nuPlan, a comprehensive real-world autonomous driving dataset and benchmark designed to evaluate ML-based planning methods in diverse, complex scenarios, aiming to accelerate the adoption of learning-based planning in autonomous vehicles.

Contribution

It provides the first large-scale real-world driving dataset and benchmark specifically for ML-based planning, including detailed scenario taxonomy and a simulation framework for evaluation.

Findings

ML-based planners show gaps in handling rare scenarios

Traditional methods outperform some ML-based approaches in safety

The dataset enables detailed analysis of planning performance

Abstract

Machine Learning (ML) has replaced traditional handcrafted methods for perception and prediction in autonomous vehicles. Yet for the equally important planning task, the adoption of ML-based techniques is slow. We present nuPlan, the world's first real-world autonomous driving dataset, and benchmark. The benchmark is designed to test the ability of ML-based planners to handle diverse driving situations and to make safe and efficient decisions. To that end, we introduce a new large-scale dataset that consists of 1282 hours of diverse driving scenarios from 4 cities (Las Vegas, Boston, Pittsburgh, and Singapore) and includes high-quality auto-labeled object tracks and traffic light data. We exhaustively mine and taxonomize common and rare driving scenarios which are used during evaluation to get fine-grained insights into the performance and characteristics of a planner. Beyond the dataset, we provide a simulation and evaluation framework that enables a planner's actions to be simulated in closed-loop to account for interactions with other traffic participants. We present a detailed analysis of numerous baselines and investigate gaps between ML-based and traditional methods. Find the nuPlan dataset and code at nuplan.org.