Bridging Predictive Uncertainty and Safe Action: Sample-Conditioned Differentiable Planning for Autonomous Driving
2026-06-02 • Robotics
Robotics
AI summaryⓘ
The authors address the challenge of driving safely in unpredictable traffic by combining future scenario predictions with planning in a way that is easy to understand and reliable. They use a method that generates many possible future traffic movements (called diffusion-generated trajectories) and directly include these in a planning system that tries to avoid risky situations. Their planner uses a special safety measure (CVaR) to be cautious about rare but dangerous events. They also improved how the traffic scene is represented using directed graphs to make predictions better and faster. Tests on real-world driving datasets show their approach improves safety, smoothness, and efficiency compared to previous methods.
Autonomous drivingDiffusion modelsTrajectory predictionMotion planningConditional Value-at-Risk (CVaR)Directed graphsUncertainty modelingSafety constraintsWaymo Open Motion datasetArgoverse 2 dataset
Authors
Chengzhen Meng, Pei Liu, Zhiyu Huang, Chen Lv, Jun Ma
Abstract
Complex, dynamic, and interactive driving environments pose significant challenges for autonomous driving, primarily due to the pervasive uncertainty of surrounding traffic. A fundamental bottleneck in current systems is the disconnect between highly expressive uncertainty modeling and interpretable, safe motion planning. In this paper, we propose a novel sample-conditioned differentiable planning framework that bridges this gap by explicitly incorporating diffusion-generated future trajectories into the optimization process. Rather than compressing predictions into a single deterministic future or relying on black-box end-to-end architectures, our approach leverages a conditional diffusion model to generate a diverse set of plausible future scenarios. Crucially, these samples are directly fed into a differentiable planner, which explicitly mitigates predictive uncertainty via an empirical Conditional Value-at-Risk (CVaR) tail-risk constraint. This allows the planner to optimize a physically interpretable trajectory that is robust to rare yet safety-critical interactions. Furthermore, we introduce a directed graph representation for scene context that yields substantial improvements in both predictive effectiveness and computational efficiency. Validated through extensive open-loop and closed-loop evaluations on the Waymo Open Motion and Argoverse 2 datasets, our framework significantly outperforms state-of-the-art baselines in safety, efficiency, and ride comfort.