• Corpus ID: 237532368

Learning Observation-Based Certifiable Safe Policy for Decentralized Multi-Robot Navigation

  title={Learning Observation-Based Certifiable Safe Policy for Decentralized Multi-Robot Navigation},
  author={Yuxiang Cui and Longzhong Lin and Xiaolong Huang and Dongkun Zhang and Yue Wang and Rong Xiong},
Safety is of great importance in multi-robot navigation problems. In this paper, we propose a control barrier function (CBF) based optimizer that ensures robot safety with both high probability and flexibility, using only sensor measurement. The optimizer takes action commands from the policy network as initial values and then provides refinement to drive the potentially dangerous ones back into safe regions. With the help of a deep transition model that predicts the evolution of surrounding… 

Figures and Tables from this paper


Where to go Next: Learning a Subgoal Recommendation Policy for Navigation in Dynamic Environments
This letter proposes to learn, via deep Reinforcement Learning (RL), an interaction-aware policy that provides long-term guidance to the local planner that is shown to substantially improve the navigation performance in terms of number of collisions as compared to prior MPC frameworks.
Fully Distributed Multi-Robot Collision Avoidance via Deep Reinforcement Learning for Safe and Efficient Navigation in Complex Scenarios
This paper presents a decentralized sensor-level collision avoidance policy for multi-robot systems, which shows promising results in practical applications and directly maps raw sensor measurements to an agent's steering commands in terms of the movement velocity.
Motion Planning Among Dynamic, Decision-Making Agents with Deep Reinforcement Learning
This work extends the previous approach to develop an algorithm that learns collision avoidance among a variety of types of dynamic agents without assuming they follow any particular behavior rules, and introduces a strategy using LSTM that enables the algorithm to use observations of an arbitrary number of other agents, instead of previous methods that have a fixed observation size.
Learning World Transition Model for Socially Aware Robot Navigation
A model-based reinforcement learning approach for robots to navigate through crowded environments by taking laser scan sequence and robot’s own state as input and outputs steering command and guiding the robots to the final targets successfully in a socially compliant manner.
MAMPS: Safe Multi-Agent Reinforcement Learning via Model Predictive Shielding
This work proposes multi-agent model predictive shielding (MAMPS), an algorithm that provably guarantees safety for an arbitrary learned policy and operates by using the learned policy as often as possible, but instead uses a backup policy in cases where it cannot guarantee the safety of the learning policy.
End-to-End Safe Reinforcement Learning through Barrier Functions for Safety-Critical Continuous Control Tasks
This work proposes a controller architecture that combines a model-free RL-based controller with model-based controllers utilizing control barrier functions (CBFs) and on-line learning of the unknown system dynamics, in order to ensure safety during learning.
Socially-Aware Multi-Agent Following with 2D Laser Scans via Deep Reinforcement Learning and Potential Field
This paper proposes a multi-agent method for an arbitrary number of robots to follow the target in a socially-aware manner using only 2D laser scans, and shows that the method works successfully in unseen dynamic environments.
Neural Certificates for Safe Control Policies
The safety means that a policy must not drive the state of the system to any unsafe region, while the goal-reaching requires the trajectory of the controlled system asymptotically converges to a goal region (a generalization of stability).
Synthesis of Control Barrier Functions Using a Supervised Machine Learning Approach
A learning framework for estimating control barrier functions from sensor data that affords system operation in unknown state space regions without compromising safety is described and theoretical safety guarantees are provided.
Crowd-Robot Interaction: Crowd-Aware Robot Navigation With Attention-Based Deep Reinforcement Learning
This work proposes to rethink pairwise interactions with a self-attention mechanism, and jointly model Human-Robot as well as Human-Human interactions in the deep reinforcement learning framework, and captures the Human- human interactions occurring in dense crowds that indirectly affects the robot’s anticipation capability.