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

2024-11-09

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.

Robotics

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: how to develop a mobile robot localization system without pre - installed infrastructure and without drift. Specifically, although the existing infrastructure - based localization methods have reliable drift - free localization capabilities, the pre - installed infrastructure has the problems of poor flexibility and high maintenance costs. Therefore, this paper proposes a mobile robot localization system without infrastructure and without drift by using environmental magnetic field (MF) information, namely IDF - MFL.

### Problem Background
1. **Limitations of Existing Technologies**
- **Drift Problem of SLAM Systems**: In closed or semi - closed environments (such as corridors, industrial warehouses, underground parking lots), SLAM systems will inevitably produce drift. Even though the drift can be corrected by loop closure and bundle adjustment techniques, the sharp change of state is unacceptable for mobile robots.
- **Infrastructure Dependence Problem**: In order to overcome the drift problem of SLAM systems, pre - installed infrastructure (such as UWB anchors, RFID tags, two - dimensional codes, etc.) is usually deployed in industrial scenarios. However, the installation and maintenance processes of these infrastructures are neither flexible nor expensive, which limits the wide application of mobile robots.

2. **Inspiration from Nature**
- Inspired by some animals in nature (such as thornbirds and lobsters) that can use local anomalies of the earth's magnetic field to sense position and direction, this paper studies the use of environmental magnetic field information to realize a mobile robot localization system without infrastructure.

### Solution
The IDF - MFL system proposed in this paper uses the magnetic field information in the environment for localization and mainly solves the following key problems:
- **No Need for Pre - installed Infrastructure**: By using the highly unique magnetic field information generated by ferromagnetic objects inherent in the environment (such as steel and reinforced concrete structures of buildings, underground pipelines, etc.), localization without pre - installed infrastructure is achieved.
- **Drift - free Localization**: A robust state estimation algorithm is proposed, which can handle non - Gaussian heavy - tailed noise (such as measurement outliers) caused by dynamic ferromagnetic objects, thus ensuring the drift - free characteristic of localization.
- **Real - time and Accuracy**: By optimizing the expected value of the magnetic field matching cost function, high - precision, reliable and real - time localization is achieved.

### Main Contributions
1. **Pure Magnetic Field Localization Method**: A pure magnetic field localization method without any auxiliary odometer information is proposed, which is suitable for various mobile robot platforms.
2. **Robust State Estimator**: A robust state estimator that can handle non - Gaussian heavy - tailed noise is designed, and the non - convex and non - continuous state estimation problem is solved by a stochastic optimization algorithm.
3. **Verification in Practical Applications**: The practical implementation of the stochastic estimator is described in detail, and the feasibility and performance of the system are verified through simulation and real - world experiments.

In summary, this paper aims to develop a mobile robot localization system without pre - installed infrastructure and without drift to solve the problems of localization drift and infrastructure dependence of existing technologies in closed or semi - closed environments.