Machine Learning-Based Position Detection Using Hall-Effect Sensor Arrays on Resource-Constrained Microcontroller

Aug 2, 2026

Upload by Skyler

This paper presents an electromagnetic levitation system that stabilizes a magnetic body using an array of electromagnets guided by a Hall-effect sensor array and TinyML-based position detection. The approach replaces optical tracking with finite-element optimized electromagnets and a microcontroller-optimized neural network that predicts the levitated object’s position from sensor data with 0.0263–0.0381 mm mean absolute error. Quantized and full-precision supervised multi-output regression models are trained on spatially sampled data and comprehensively benchmarked. Results show stable 850–1000 Hz control frequencies matching optical systems while removing cost and complexity, enabling real-time on-board position detection and current calculation and validating sub-30 µm accuracy on standard microcontrollers.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages16

Reading Time40 min

AccessFree to download

Introduction

Magnetic levitation advantages and challenges

Applications of Maglev in different domains

Physical principles and need for dynamic feedback

Article Machine Learning-Based Position Detection Using Hall-Effect Sensor Arrays on Resource-Constrained Microcontroller Zalán Németh 1, Chan Hwang See 1, *, Keng Goh 1, Arfan Ghani 2, Simeon Keates 3 and Raed A. Abd-Alhameed 4,5 Academic Editor: Daniel Ramos Received: 10 September 2025 Revised: 3 October 2025 Accepted: 16 October 2025 Published: 18 October 2025 Citation: Németh, Z.; See, C.H.; Goh, K.; Ghani, A.; Keates, S.; A. AbdAlhameed, R. Machine LearningBased Position Detection Using Hall-Effect Sensor Arrays on ResourceConstrained Microcontroller. Sensors 2025, 25, 6444. [https://doi.org/](https://doi.org/)[ ](https://doi.org/)[10.3390/s25206444](10.3390/s25206444) Copyright: © 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ([https://creativecommons.org/](https://creativecommons.org/)[ ](https://creativecommons.org/)[licenses/by/4.0/](licenses/by/4.0/)) . 1 School of Computing, Engineering and the Built Environment, Edinburgh Napier University, Edinburgh EH10 5DT, UK; [zal.nemeth@gmail.com](zal.nemeth@gmail.com) (Z.N.); [k.goh@napier.ac.uk](k.goh@napier.ac.uk) (K.G.) 2 Department of Computer Science and Engineering, American University of Ras al Khaimah, Ras al Khaimah 72603, United Arab Emirates; [arfan.ghani@aurak.ac.ae](arfan.ghani@aurak.ac.ae) 3 Centre for Future Technologies, University of Chichester, Chichester PO19 6PE, UK; [s.keates@chi.ac.uk](s.keates@chi.ac.uk) 4 Faculty of Engineering and Digital Technologies, University of Bradford, Bradford BD7 1DP, UK; [r.a.a.abd@bradford.ac.uk](r.a.a.abd@bradford.ac.uk) 5 Department of Information and Communication Engineering, Al-Farqadein University College, Basrah 651004, Iraq * Correspondence: [c.see@napier.ac.uk](c.see@napier.ac.uk) Abstract This paper presents an electromagnetic levitation system that stabilizes a magnetic body using an array of electromagnets controlled by a Hall-effect sensor array and TinyMLbased position detection. Departing from conventional optical tracking methods, the proposed design combines finite-element-optimized electromagnets with a microcontrolleroptimized neural network that processes sensor data to predict the levitated object’s position with 0.0263–0.0381 mm mean absolute error. The system employs both quantized and full-precision implementations of a supervised multi-output regression model trained on spatially sampled data (40 × 40 × 15 mm volume at 5 mm intervals) . Comprehensive benchmarking demonstrates stable operation at 850–1000 Hz control frequencies, matching optical systems’ performance while eliminating their cost and complexity. The integrated solution performs real-time position detection and current calculation entirely on-board, requiring no external tracking devices or high-performance computing. By achieving sub 30 µm accuracy with standard microcontrollers and minimal hardware, this work validates machine learning as a viable alternative to optical position detection in magnetic levitation systems, reducing implementation barriers for research and industrial applications. The complete system design, including electromagnetic array characterization, neural network architecture selection, and real-time implementation challenges, is presented alongside performance comparisons with conventional approaches. Keywords: machine learning; Hall-effect sensor array; electromagnetic levitation system; microcontroller; TinyML 1. Introduction Magnetic levitation (Maglev) technology has garnered widespread recognition for its transformative capabilities in high-precision positioning systems, with applications ranging from industrial automation and transportation to semiconductor manufacturing [1] . The core advantages of Maglev systems include the elimination of mechanical wear, friction, backlash, and vibration, which are prevalent in traditiona

FAQs

What positioning method does the paper propose instead of optical tracking?

It proposes Hall-effect sensor arrays with TinyML-based position detection using an on-board microcontroller-optimized neural network.

How accurate is the proposed position detection system?

The system achieves a mean absolute error in the range of 0.0263–0.0381 mm and demonstrates sub-30 µm accuracy with standard microcontrollers.

What control performance is reported for the levitation system?

Benchmarking demonstrates stable operation at 850–1000 Hz control frequencies while maintaining performance comparable to optical systems.

Aug 2, 2026

Upload by Skyler