Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Snake-Inspired Mobile Robot Positioning with Hybrid Learning

Aviad Etzion, Nadav Cohen, Orzion Levy, Zeev Yampolsky, Itzik Klein

TL;DR

MoRPINet addresses the challenge of drift in pure inertial navigation for mobile robots by introducing a snake-inspired serpentine movement and a neural distance estimator (D-Net) to produce reliable 2D positions from IMU data. The framework couples D-Net distance regression with Madgwick-based heading to perform dead-reckoning updates at a higher rate than prior peak-to-peak methods (MoRPI), demonstrating a ~30% reduction in PRMSE compared with MoRPI and strong distance estimation accuracy (DRMSE ≈ 0.022 m). Field experiments with multiple IMUs and RTK-GNSS ground truth show MoRPINet outperforms INS by a wide margin, while offering a practical, edge-friendly model with publicly available data and code. The work highlights serpentine dynamics as a viable strategy to enrich inertial signals and improve pure inertial odometry for mobile robots in challenging environments.

Abstract

Mobile robots are used in various fields, from deliveries to search and rescue applications. Different types of sensors are mounted on the robot to provide accurate navigation and, thus, allow successful completion of its task. In real-world scenarios, due to environmental constraints, the robot frequently relies only on its inertial sensors. Therefore, due to noises and other error terms associated with the inertial readings, the navigation solution drifts in time. To mitigate the inertial solution drift, we propose the MoRPINet framework consisting of a neural network to regress the robot's travelled distance. To this end, we require the mobile robot to maneuver in a snake-like slithering motion to encourage nonlinear behavior. MoRPINet was evaluated using a dataset of 290 minutes of inertial recordings during field experiments and showed an improvement of 33% in the positioning error over other state-of-the-art methods for pure inertial navigation.

Snake-Inspired Mobile Robot Positioning with Hybrid Learning

TL;DR

MoRPINet addresses the challenge of drift in pure inertial navigation for mobile robots by introducing a snake-inspired serpentine movement and a neural distance estimator (D-Net) to produce reliable 2D positions from IMU data. The framework couples D-Net distance regression with Madgwick-based heading to perform dead-reckoning updates at a higher rate than prior peak-to-peak methods (MoRPI), demonstrating a ~30% reduction in PRMSE compared with MoRPI and strong distance estimation accuracy (DRMSE ≈ 0.022 m). Field experiments with multiple IMUs and RTK-GNSS ground truth show MoRPINet outperforms INS by a wide margin, while offering a practical, edge-friendly model with publicly available data and code. The work highlights serpentine dynamics as a viable strategy to enrich inertial signals and improve pure inertial odometry for mobile robots in challenging environments.

Abstract

Mobile robots are used in various fields, from deliveries to search and rescue applications. Different types of sensors are mounted on the robot to provide accurate navigation and, thus, allow successful completion of its task. In real-world scenarios, due to environmental constraints, the robot frequently relies only on its inertial sensors. Therefore, due to noises and other error terms associated with the inertial readings, the navigation solution drifts in time. To mitigate the inertial solution drift, we propose the MoRPINet framework consisting of a neural network to regress the robot's travelled distance. To this end, we require the mobile robot to maneuver in a snake-like slithering motion to encourage nonlinear behavior. MoRPINet was evaluated using a dataset of 290 minutes of inertial recordings during field experiments and showed an improvement of 33% in the positioning error over other state-of-the-art methods for pure inertial navigation.

Paper Structure

This paper contains 19 sections, 16 equations, 6 figures, 7 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Model-Based Approaches
  3. Inertial Navigation System
  4. MoRPI
  5. MoRPINet Framework
  6. D-Net: Distance Estimation Network
  7. MoRPINet Training Process
  8. Madgwick Filter
  9. Summary
  10. Experiment Setup and Dataset
  11. Experiment Setup
  12. Dataset
  13. Analysis and Results
  14. Evaluation Metrics
  15. Field Experiment Results
  16. ...and 4 more sections

Figures (6)

  • Figure 1: MoRPINet is an inertial positioning approach for mobile robots.
  • Figure 2: D-Net: Our inertial-based distance estimation network architecture.
  • Figure 3: Our RC car mounted with (1) GNSS RTK receiver and (2) five IMU sensors.
  • Figure 4: Example of four different trajectories. The mobile robot is moving in a snake-like slithering motion.
  • Figure 6: An illustration of ground-truth trajectory (in red). The distance (in green) and the heading angles are calculated using \ref{['eq:gt_dist']}-\ref{['eq:gt_angs']}.
  • ...and 1 more figures