Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Distributed Perception Aware Safe Leader Follower System via Control Barrier Methods

Richie R. Suganda, Tony Tran, Miao Pan, Lei Fan, Qin Lin, Bin Hu

TL;DR

This work tackles distributed leader–follower formation under limited body-fixed camera FOV by embedding perception constraints into a safety-preserving control framework. A distributed CBF-QP balances formation objectives with forward invariance of FOV-based safety sets, ensuring the leader remains visible to each follower. To support real-time perception under diverse environments, the authors introduce a temporal-filtered DNN for estimating bearing $\phi_i$ and a double bounding-box method for distance $L_i$, with domain randomization and a linear temporal filter to boost robustness. Validation in Gazebo demonstrates robust formation control across environments, highlighting the practical viability of perception-aware safety mediation in GPS-denied multi-agent systems. The approach offers a principled path to perception-constrained coordination with explicit safety guarantees and data-driven perception modules.

Abstract

This paper addresses a distributed leader-follower formation control problem for a group of agents, each using a body-fixed camera with a limited field of view (FOV) for state estimation. The main challenge arises from the need to coordinate the agents' movements with their cameras' FOV to maintain visibility of the leader for accurate and reliable state estimation. To address this challenge, we propose a novel perception-aware distributed leader-follower safe control scheme that incorporates FOV limits as state constraints. A Control Barrier Function (CBF) based quadratic program is employed to ensure the forward invariance of a safety set defined by these constraints. Furthermore, new neural network based and double bounding boxes based estimators, combined with temporal filters, are developed to estimate system states directly from real-time image data, providing consistent performance across various environments. Comparison results in the Gazebo simulator demonstrate the effectiveness and robustness of the proposed framework in two distinct environments.

Distributed Perception Aware Safe Leader Follower System via Control Barrier Methods

TL;DR

This work tackles distributed leader–follower formation under limited body-fixed camera FOV by embedding perception constraints into a safety-preserving control framework. A distributed CBF-QP balances formation objectives with forward invariance of FOV-based safety sets, ensuring the leader remains visible to each follower. To support real-time perception under diverse environments, the authors introduce a temporal-filtered DNN for estimating bearing and a double bounding-box method for distance , with domain randomization and a linear temporal filter to boost robustness. Validation in Gazebo demonstrates robust formation control across environments, highlighting the practical viability of perception-aware safety mediation in GPS-denied multi-agent systems. The approach offers a principled path to perception-constrained coordination with explicit safety guarantees and data-driven perception modules.

Abstract

This paper addresses a distributed leader-follower formation control problem for a group of agents, each using a body-fixed camera with a limited field of view (FOV) for state estimation. The main challenge arises from the need to coordinate the agents' movements with their cameras' FOV to maintain visibility of the leader for accurate and reliable state estimation. To address this challenge, we propose a novel perception-aware distributed leader-follower safe control scheme that incorporates FOV limits as state constraints. A Control Barrier Function (CBF) based quadratic program is employed to ensure the forward invariance of a safety set defined by these constraints. Furthermore, new neural network based and double bounding boxes based estimators, combined with temporal filters, are developed to estimate system states directly from real-time image data, providing consistent performance across various environments. Comparison results in the Gazebo simulator demonstrate the effectiveness and robustness of the proposed framework in two distinct environments.
Paper Structure (14 sections, 14 equations, 10 figures)

This paper contains 14 sections, 14 equations, 10 figures.

Table of Contents

  1. Introduction
  2. Preliminaries
  3. System Framework and Problem Formulation
  4. Distributed Perception-aware Safe Leader-follower Formation Control
  5. Leader-follower System State Estimation
  6. Estimating $\phi_i$ Using a Temporal Filter-Based DNN
  7. Training Data Collection
  8. Data Labeling
  9. DNN with a Linear Temporal Filter
  10. Distance $L_i$ Estimation via Double Bounding Boxes
  11. Experiment Results
  12. Simulation Setup
  13. Simulation results
  14. CONCLUSIONS

Figures (10)

  • Figure 1: Perception-aware Leader-follower (LF) System with Limited FOV. LF formations are composed of robots equipped with onboard body-fixed camera that have limited FOV. To maintain safe and reliable formations, the leader must remain within the follower's FOV.
  • Figure 2: Leader-follower formation diagram.
  • Figure 3: Distributed Perception-aware Safe Leader-follower Control and Estimation Architecture.
  • Figure 4: Data Labeling. Image data labeling example with six classes.
  • Figure 5: Comparison Results of Angle Estimation Error with and without Linear Temporal Filter.
  • ...and 5 more figures

Theorems & Definitions (1)

  • Definition 2.1