UncAD: Towards Safe End-to-end Autonomous Driving via Online Map Uncertainty

2025-04-17

AI Summary

To address perception bias and planning safety risks arising from overreliance on deterministic online maps in end-to-end autonomous driving, this paper proposes, for the first time, an online map uncertainty-driven paradigm. Methodologically, we design a deep learning model to estimate pixel-wise uncertainty of online maps and integrate it into multimodal motion prediction (via uncertainty-weighted fusion) and trajectory selection (via joint modeling of uncertainty and collision risk), enabling uncertainty-aware collision avoidance. Our core contributions are: (1) the first end-to-end framework that explicitly models and propagates map uncertainty across the entire perception–prediction–planning pipeline; and (2) an uncertainty-guided multimodal prediction mechanism coupled with a unified uncertainty–collision risk assessment for trajectory evaluation. Integrated with state-of-the-art (SOTA) methods on nuScenes, our approach incurs only a 1.9% parameter overhead while reducing collision rate by up to 26% and drivable-area conflict rate by up to 42%.

Technology Category

• Reasoning under Uncertainty: Uncertainty Representations
• Planning, Routing, and Scheduling: Planning under Uncertainty
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

End-to-end autonomous driving aims to produce planning trajectories from raw sensors directly. Currently, most approaches integrate perception, prediction, and planning modules into a fully differentiable network, promising great scalability. However, these methods typically rely on deterministic modeling of online maps in the perception module for guiding or constraining vehicle planning, which may incorporate erroneous perception information and further compromise planning safety. To address this issue, we delve into the importance of online map uncertainty for enhancing autonomous driving safety and propose a novel paradigm named UncAD. Specifically, UncAD first estimates the uncertainty of the online map in the perception module. It then leverages the uncertainty to guide motion prediction and planning modules to produce multi-modal trajectories. Finally, to achieve safer autonomous driving, UncAD proposes an uncertainty-collision-aware planning selection strategy according to the online map uncertainty to evaluate and select the best trajectory. In this study, we incorporate UncAD into various state-of-the-art (SOTA) end-to-end methods. Experiments on the nuScenes dataset show that integrating UncAD, with only a 1.9% increase in parameters, can reduce collision rates by up to 26% and drivable area conflict rate by up to 42%. Codes, pre-trained models, and demo videos can be accessed at https://github.com/pengxuanyang/UncAD.

Problem

• Addresses unsafe planning from deterministic online maps in autonomous driving
• Proposes uncertainty-aware perception for safer trajectory prediction and planning
• Reduces collisions and conflicts via uncertainty-guided trajectory selection

Innovation

• Estimates online map uncertainty in perception
• Leverages uncertainty for multi-modal trajectory planning
• Uses uncertainty-collision-aware trajectory selection strategy