Reactive Collision Avoidance for Safe Agile Navigation

2024-09-18

arXiv.org

AI Summary

This work addresses the challenges of real-time obstacle avoidance, high control latency, and the trade-off between safety and agility for autonomous micro aerial vehicles operating in unknown, dynamic environments. We propose a map-free, parameter-free, end-to-end closed-loop navigation framework. Methodologically, it introduces the first deep integration of nonlinear model predictive control (NMPC) with adaptive control barrier functions (CBFs), augmented by a lightweight RGB-D temporal neural network for depth estimation and a minimum-time-to-collision–driven threat-prioritization mechanism; dynamic heuristic optimization further enables online balancing of safety and agility. Experiments across diverse indoor and outdoor dynamic scenarios demonstrate zero mapping overhead and zero manual tuning, achieving high-speed flight with zero collisions, significantly reduced response latency, and minimized over-conservative constraints—establishing a new paradigm for real-time reactive navigation in complex dynamic environments.

Technology Category

• Multiagent Systems: Adversarial Agents
• Intelligent Robots: State Estimation
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Reactive collision avoidance is essential for agile robots navigating complex and dynamic environments, enabling real-time obstacle response. However, this task is inherently challenging because it requires a tight integration of perception, planning, and control, which traditional methods often handle separately, resulting in compounded errors and delays. This paper introduces a novel approach that unifies these tasks into a single reactive framework using solely onboard sensing and computing. Our method combines nonlinear model predictive control with adaptive control barrier functions, directly linking perception-driven constraints to real-time planning and control. Constraints are determined by using a neural network to refine noisy RGB-D data, enhancing depth accuracy, and selecting points with the minimum time-to-collision to prioritize the most immediate threats. To maintain a balance between safety and agility, a heuristic dynamically adjusts the optimization process, preventing overconstraints in real time. Extensive experiments with an agile quadrotor demonstrate effective collision avoidance across diverse indoor and outdoor environments, without requiring environment-specific tuning or explicit mapping.

Problem

• Reactive collision avoidance for agile robots in dynamic environments
• Integration of perception, planning, and control to reduce errors
• Balancing safety and agility in real-time obstacle response

Innovation

• Unifies perception, planning, control in reactive framework
• Combines nonlinear MPC with adaptive barrier functions
• Refines RGB-D data via neural network for accuracy