Research Analyzer
← Back ICRA 2026

SPOT: Spatio-Temporal Trajectory Planning for UAVs in Unknown Dynamic Environments

Astik Srivastava, Bhanu Teja Bitla, Thomas J Chackenkulam, Antony Thomas, Madhava Krishna

PDF

AI summary

Key figure (auto-extracted from paper)
SPOT enables robust, mapless UAV navigation in unknown dynamic environments by combining spatio-temporal path planning with reactive backup strategies to prevent deadlocks and collisions.
UAV navigation dynamic obstacle avoidance spatio-temporal planning mapless navigation backup planning RRT*

Problem

Current UAV motion planners degrade in cluttered, dynamic settings due to reliance on map fusion, ground-truth data, or a lack of fallback strategies for deadlock situations.

Approach

The framework uses vision-based object tracking to classify static and dynamic obstacles, plans spatio-temporal paths with an RRT* variant, constructs time-aware safe flight corridors, and triggers a reactive backup planner when direct paths are blocked.

Key results

  • Novel spatio-temporal RRT* algorithm for mapless navigation
  • Reactive backup planning module to resolve deadlocks
  • Successful real-world and simulation validation with up to 30 dynamic obstacles
  • Open-source code release for reproducibility

Why it matters

Provides a scalable, onboard-computable solution for safe autonomous UAV operation in unpredictable, real-world environments.

Abstract

We address the problem of reactive motion plan- ning for quadrotors operating in unknown environments with dynamic obstacles. Our approach leverages a 4-dimensional spatio-temporal planner, integrated with vision-based Safe Flight Corridor (SFC) generation and trajectory optimization. Unlike prior methods that rely on map fusion, our framework is mapless, enabling collision avoidance directly from perception while reducing computational overhead. Dynamic obstacles are detected and tracked using a vision-based object segmentation and tracking pipeline, allowing robust classification of static ver- sus dynamic elements in the scene. To further enhance robustness, we introduce a backup planning module that reactively avoids dynamic obstacles when no direct path to the goal is available, mitigating the risk of collisions during deadlock situations. We validate our method extensively in both simulation and real-world hardware experiments, and benchmark it against state-of-the- art approaches, showing significant advantages for reactive UAV navigation in dynamic, unknown environments.

Index terms

Aerial Systems: Perception and Autonomy Collision Avoidance Motion and Path Planning

Related papers