Get Started
Home
Topics
Search
Library
Research questionHow can hybrid autonomous-driving planners align trajectory generation and refinement when their objectives differ?Hybrid planners may generate and refine trajectories with separate objectives, causing the refined plan to diverge from the original and produce suboptimal driving behavior. The central difficulty is coordinating collision avoidance, drivable-area compliance, comfort, and progress across both stages.
AI
Diffusion Models
Evaluation & Benchmarks
Robotics
Latest papersRecent research connected to this question, newest first.DiffuSearch: How Hybrid Trajectory Planning Benefits from Aligned Objectives in Diffusion and Action SpaceThe source studies a two-stage planner that uses guided diffusion for joint trajectory generation and Monte Carlo Tree Search for local action-space refinement, with shared driving objectives. Evidence comes from nuPlan and interPlan reactive closed-loop benchmarks and ablation studies.research paper · Sep 2, 2026
Related questions
How can autonomous-driving planners be stress-tested in realistic closed-loop scenarios that expose failures missed by nominal benchmarks?How can LLM-based vehicle negotiation avoid premature agreement when multiple driving decisions are viable?How can autonomous-driving scenario generators reliably induce collisions at a requested region of the target vehicle?How can spacecraft rendezvous planners learn uncertainty-aware dynamics from offline trajectories while generating enough training data?