Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

IDF-MFL: Infrastructure-free and Drift-free Magnetic Field Localization for Mobile Robot

Hongming Shen, Zhenyu Wu, Wei Wang, Qiyang Lyu, Huiqin Zhou, Danwei Wang

TL;DR

The results demonstrate that the proposed method can achieve high-accuracy, reliable, and real-time localization without any pre-installed infrastructures.

Abstract

In recent years, infrastructure-based localization methods have achieved significant progress thanks to their reliable and drift-free localization capability. However, the pre-installed infrastructures suffer from inflexibilities and high maintenance costs. This poses an interesting problem of how to develop a drift-free localization system without using the pre-installed infrastructures. In this paper, an infrastructure-free and drift-free localization system is proposed using the ambient magnetic field (MF) information, namely IDF-MFL. IDF-MFL is infrastructure-free thanks to the high distinctiveness of the ambient MF information produced by inherent ferromagnetic objects in the environment, such as steel and reinforced concrete structures of buildings, and underground pipelines. The MF-based localization problem is defined as a stochastic optimization problem with the consideration of the non-Gaussian heavy-tailed noise introduced by MF measurement outliers (caused by dynamic ferromagnetic objects), and an outlier-robust state estimation algorithm is derived to find the optimal distribution of robot state that makes the expectation of MF matching cost achieves its lower bound. The proposed method is evaluated in multiple scenarios, including experiments on high-fidelity simulation, and real-world environments. The results demonstrate that the proposed method can achieve high-accuracy, reliable, and real-time localization without any pre-installed infrastructures.

IDF-MFL: Infrastructure-free and Drift-free Magnetic Field Localization for Mobile Robot

TL;DR

The results demonstrate that the proposed method can achieve high-accuracy, reliable, and real-time localization without any pre-installed infrastructures.

Abstract

In recent years, infrastructure-based localization methods have achieved significant progress thanks to their reliable and drift-free localization capability. However, the pre-installed infrastructures suffer from inflexibilities and high maintenance costs. This poses an interesting problem of how to develop a drift-free localization system without using the pre-installed infrastructures. In this paper, an infrastructure-free and drift-free localization system is proposed using the ambient magnetic field (MF) information, namely IDF-MFL. IDF-MFL is infrastructure-free thanks to the high distinctiveness of the ambient MF information produced by inherent ferromagnetic objects in the environment, such as steel and reinforced concrete structures of buildings, and underground pipelines. The MF-based localization problem is defined as a stochastic optimization problem with the consideration of the non-Gaussian heavy-tailed noise introduced by MF measurement outliers (caused by dynamic ferromagnetic objects), and an outlier-robust state estimation algorithm is derived to find the optimal distribution of robot state that makes the expectation of MF matching cost achieves its lower bound. The proposed method is evaluated in multiple scenarios, including experiments on high-fidelity simulation, and real-world environments. The results demonstrate that the proposed method can achieve high-accuracy, reliable, and real-time localization without any pre-installed infrastructures.

Paper Structure

This paper contains 15 sections, 1 theorem, 17 equations, 5 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Works
  3. Motivation and Contributions
  4. System Overview
  5. Problem formulation
  6. Robust Estimation via Stochastic Optimization
  7. Practical Implementation for Magnetic Field-based State Estimation
  8. Reducing the Sampling Dimension on Manifolds
  9. Improve the Numerical Stability: Cost Shifting
  10. Results
  11. Comparison Baseline
  12. Accuracy Evaluations
  13. Robustness Evaluations
  14. Running Time Evaluations
  15. Conclusion

Key Result

Theorem 1

The lower bound of $\mathbb{E}_{\mathcal{Q}^*}\left[S\left(\mathbf{x}_{t_k}\right)\right]$ can be defined as: where $\lambda \in \mathbb{R}^+$ and ${\mathbb{D}}\left( {{{\@fontswitch\mathcal{Q}}^*}||{\@fontswitch\mathcal{Q}}} \right)$ denotes the KL divergence between distributions ${\@fontswitch\mathcal{Q}}^*$ and ${\@fontswitch\mathcal{Q}}$.

Figures (5)

  • Figure 1: System overview of IDF-MFL. As noted in Remark \ref{['rem: pure MF']}, the external odometry is only used in offline MF map construction phase.
  • Figure 2: High-fidelity physical simulation environment and the mobile robot platform used in real-world experiments.
  • Figure 3: MF distribution and trajectory estimation results. In each map, color designates the MF integrated intensity ($l^2-$norm) according to the scale shown in each map. Trajectories estimated by the proposed method and ground truth are noted with white and yellow, respectively.
  • Figure 4: Comparations of localizability for investigated MF-based localization methods.
  • Figure 5: Velocity estimated from the proposed method and ground truth under the Corridor sequence. It is worth noting that velocity can change drastically when loop closure occurs. The velocity ground truth is generated by FAST-LIO2FASTLIO2.

Theorems & Definitions (4)

  • Remark 1
  • Theorem 1
  • proof
  • Remark 2