AVOCADO: Adaptive Optimal Collision Avoidance driven by Opinion

2024-06-29

IEEE Transactions on robotics

AI Summary

Real-time collision avoidance for omnidirectional robots in multi-agent environments remains challenging due to uncertainty in neighboring agents’ cooperation levels. Method: This paper proposes an adaptive collision avoidance framework built upon the Velocity Obstacle (VO) paradigm. It introduces nonlinear opinion dynamics—novel in robotic navigation—to estimate neighbors’ cooperative intent online using only local sensor observations. Based on these estimates, an adaptive optimal controller is formulated, and a deadlock-resolution mechanism is designed leveraging symmetry in opinion evolution. The approach integrates geometric avoidance, real-time sensor fusion, and adaptive control. Contribution/Results: Extensive experiments demonstrate that the method significantly improves collision avoidance success rates, reduces time-to-goal, and maintains computational efficiency in mixed cooperative/non-cooperative and high-density human-robot coexistence scenarios.

Technology Category

• Intelligent Robots: Multi-Robot Systems
• Multiagent Systems: Adversarial Agents
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems

Abstract

We present AVOCADO (AdaptiVe Optimal Collision Avoidance Driven by Opinion), a novel navigation approach to address holonomic robot collision avoidance when the degree of cooperation of the other agents in the environment is unknown. AVOCADO departs from a Velocity Obstacle's formulation akin to the Optimal Reciprocal Collision Avoidance method. However, instead of assuming reciprocity, AVOCADO poses an adaptive control problem that aims at adapting in real-time to the cooperation degree of other robots and agents. Adaptation is achieved through a novel nonlinear opinion dynamics design that relies solely on sensor observations. As a by-product, based on the nonlinear opinion dynamics, we propose a novel method to avoid the deadlocks under geometrical symmetries among robots and agents. Extensive numerical simulations show that AVOCADO surpasses existing geometrical, learning and planning-based approaches in mixed cooperative/non-cooperative navigation environments in terms of success rate, time to goal and computational time. In addition, we conduct multiple real experiments that verify that AVOCADO is able to avoid collisions in environments crowded with other robots and humans.

Problem

• Addresses holonomic robot collision avoidance with unknown cooperation levels.
• Adapts in real-time to varying cooperation degrees of other agents.
• Avoids deadlocks under geometrical symmetries among robots and agents.

Innovation

• Adaptive control for unknown cooperation degrees
• Nonlinear opinion dynamics for real-time adaptation
• Deadlock avoidance under geometrical symmetries