ChronoForest: Closed-Loop Multi-Tree Diffusion Planning for Efficient Bridge Search and Route Composition

2026-06-04

AI Summary

This work addresses the challenge of efficiently planning long-horizon navigation routes that satisfy both goal and waypoint constraints using only short-horizon offline trajectories. The authors propose ChronoForest, the first approach to integrate a closed-loop multi-tree diffusion mechanism with online replanning. It generates candidate paths via anchor-chained tree diffusion, verifies long-range anchor connectivity through local bridging search, and dynamically reorders waypoints. By combining time-distance guidance, multi-tree online coordination, and iterative route resampling, ChronoForest achieves success rates of 99.8%, 99.3%, and 99.5% on medium, large, and huge tasks in OGBench AntMaze-Stitch, respectively—surpassing the baseline by 34.5 percentage points on huge-scale tasks. It also significantly outperforms exhaustive planning on Hamiltonian benchmarks, delivering high path quality with low computational overhead.

Abstract

How can we plan long-horizon routes that reach designated goals, visit required waypoints, and remain short when only short-horizon offline trajectories are available? This problem matters in offline navigation because collecting sufficiently rich long-horizon data is difficult, yet real agents must still solve long-range tasks with route-level efficiency rather than mere feasibility. The difficulty is twofold: at the microscopic level, composing many short-horizon segments creates a trade-off between search cost and path quality, while at the macroscopic level, waypoint ordering requires comparing pairwise travel costs among start, goal, and waypoint anchors that are unknown before planning and increasingly unreliable when estimated only from long-range temporal distance. In this paper, we propose ChronoForest, a closed-loop planning system that couples local bridge search and online route re-solving through an anchor-chaining tree diffusion planner and an online multi-tree orchestrator. ChronoForest uses temporal distance for short-range guidance and node evaluation, while using search-time bridge evidence to validate long-range anchor connectivity and repeatedly re-solve the route. On OGBench AntMaze-Stitch, ChronoForest achieves 99.8%, 99.3%, and 99.5% success on the medium, large, and giant splits and improves giant-stitch success by up to 34.5 points over prior reported diffusion-based results. On Hamiltonian route-composition benchmarks, online re-solving corrects poor temporal orderings and improves route quality while remaining substantially cheaper than exhaustive planning.

Problem

• long-horizon planning
• offline navigation
• route composition
• waypoint ordering
• bridge search

Innovation

• closed-loop planning
• tree diffusion
• bridge search
• route composition
• temporal distance