Papers
Topics
Authors
Recent
Assistant
AI Research Assistant
Well-researched responses based on relevant abstracts and paper content.
Custom Instructions Pro
Preferences or requirements that you'd like Emergent Mind to consider when generating responses.
Gemini 2.5 Flash
Gemini 2.5 Flash 168 tok/s
Gemini 2.5 Pro 49 tok/s Pro
GPT-5 Medium 37 tok/s Pro
GPT-5 High 34 tok/s Pro
GPT-4o 99 tok/s Pro
Kimi K2 214 tok/s Pro
GPT OSS 120B 466 tok/s Pro
Claude Sonnet 4.5 37 tok/s Pro
2000 character limit reached

Risk-aware Trajectory Sampling for Quadrotor Obstacle Avoidance in Dynamic Environments (2201.06645v2)

Published 17 Jan 2022 in cs.RO

Abstract: Obstacle avoidance of quadrotors in dynamic environments is still a very open problem. Current works commonly leverage traditional static maps to represent static obstacles and the detection and tracking of moving objects (DATMO) method to model dynamic obstacles separately. The detection module requires pre-training, and the dynamic obstacles can only be modeled with certain shapes, such as cylinders or ellipsoids. This work utilizes the dual-structure particle-based (DSP) dynamic occupancy map to represent the arbitrary-shaped static obstacles and dynamic obstacles simultaneously, and proposes an efficient risk-aware sampling-based local trajectory planner to realize safe flights in this map. The trajectory is planned by sampling motion primitives generated in the state space. Each motion primitive is divided into two phases: a short-term phase with a strict risk limitation and a relatively long-term phase designed to avoid high-risk regions. The risk is evaluated with the predicted particle-form future occupancy status, considering the time dimension. With an approach to split from and merge to an arbitrary global trajectory, the planner can also be used in the tasks with preplanned global trajectories. Comparison experiments show that the obstacle avoidance system composed of the DSP map and our planner performs the best in dynamic environments. In real-world tests, our quadrotor reaches a speed of 6 m/s with the motion capture system and 2.5 m/s with everything running on a low-price single-board computer.

Citations (17)

Summary

We haven't generated a summary for this paper yet.

Lightbulb Streamline Icon: https://streamlinehq.com

Continue Learning

We haven't generated follow-up questions for this paper yet.

List To Do Tasks Checklist Streamline Icon: https://streamlinehq.com

Collections

Sign up for free to add this paper to one or more collections.

Don't miss out on important new AI/ML research

See which papers are being discussed right now on X, Reddit, and more:

“Emergent Mind helps me see which AI papers have caught fire online.”

Philip

Philip

Creator, AI Explained on YouTube