Application-Based Production and Testing of a Core-Sheath Fiber Strain Sensor for Wearable Electronics: Feasibility Study of Using the Sensors in Measuring Tri-Axial Trunk Motion Angles

Ahmad Rezaei, Tyler J. Cuthbert, Mohsen Gholami, Carlo MenonSensors 2019 · OpenReview

TL;DR

Researchers have developed a novel, low-cost core-sheath fiber strain sensor using readily available materials. This sensor, integrated into a smart sleeveless shirt, accurately measures complex 3D trunk movements, offering a promising alternative to bulky motion capture systems for real-time monitoring and rehabilitation.

What Problem Does This Paper Address?

Accurately tracking human body movements in real-time is crucial for applications ranging from sports performance analysis and rehabilitation to ergonomic assessments and virtual reality. While existing technologies like optical motion capture and Inertial Measurement Units (IMUs) exist, they often suffer from limitations such as high cost, bulky hardware, restricted mobility, and susceptibility to environmental interference (e.g., magnetic fields for IMUs). Wearable electronics, particularly those integrated into textiles, offer a promising solution for unobtrusive and comfortable motion tracking. However, developing reliable, accurate, and cost-effective strain sensors that can capture complex, multi-dimensional movements remains a significant challenge. This paper aims to bridge this gap by presenting a new fiber strain sensor and demonstrating its feasibility in a smart garment for measuring tri-axial trunk motion.

Methods

The core of this work lies in the development and characterization of a core-sheath fiber strain sensor. The sensor is fabricated using a simple one-step dip-coating process, making it accessible and cost-effective. The key innovation is the combination of a supportive, low-hysteresis elastic core (either Hytrel 3078 or Polyether-urethane urea - PEU) with a conductive piezoresistive sheath (Hytrel 3078 mixed with carbon black). This design aims to mitigate common issues with piezoresistive sensors, such as signal drift and stiffening, by allowing the sheath to return to its original length within the elastic region of the core.

Key methodological steps include:

  • Material Optimization: Investigating different core materials and varying the concentration of carbon black in the sheath to determine the optimal composition for piezoresistivity and mechanical properties.
  • Sensor Conditioning: Developing a conditioning process (repeated stretching) to stabilize sensor performance and eliminate first-cycle hysteresis and signal drift.
  • Random Wave Pattern Testing: Simulating real-world usage by subjecting the sensors to a rigorous 4-hour random wave pattern strain test to assess consistency and accuracy over extended periods.
  • Machine Learning Integration: Employing a Random Forest Regressor (RFMLA) algorithm to process the raw sensor data and predict strain, significantly improving accuracy beyond simple normalization.
  • Smart Garment Development: Integrating 18 optimized fiber strain sensors into a sleeveless shirt to capture trunk movements.
  • Kinematic Data Collection: Using a Vicon motion capture system as a ground truth for trunk angles during controlled movement studies.

Datasets

The study utilized a custom-generated dataset for training and evaluation.

  • Random Wave Pattern Data: Collected over 4 hours from sensors subjected to random wave patterns simulating typical prototype device usage. This dataset was used to evaluate sensor performance and train the RFMLA model.
  • Human Motion Capture Data: Collected from 12 healthy male participants performing a variety of uniaxial and multiaxial trunk movements. This data, captured by a Vicon motion capture system, served as the ground truth for evaluating the performance of the smart shirt system.

Key Insights and Ideas

  • Core-Sheath Design Advantage: The elastic core effectively supports the piezoresistive sheath, preventing excessive plastic deformation and reducing signal drift, a common problem in simpler piezoresistive sensors.
  • Conditioning is Crucial: A simple conditioning process (straining above the working range) significantly improves sensor stability and eliminates hysteresis, making the sensors more reliable for long-term use.
  • Machine Learning is a Game-Changer: Raw sensor data, when processed by a Random Forest Regressor, yields significantly higher accuracy than traditional normalization methods, demonstrating the power of ML in extracting meaningful information from noisy sensor signals.
  • Textile Integration is Feasible: The developed fiber sensors can be seamlessly integrated into textiles using simple stitching, paving the way for comfortable and unobtrusive wearable sensing.
  • Trunk Motion Tracking Potential: The smart shirt system demonstrates high accuracy in measuring complex 3D trunk movements, outperforming some IMU-based systems, especially in environments prone to magnetic interference.

Key Results

  • Sensor Performance: The optimized core-sheath sensors achieved a Normalized Root Mean Squared Error (NRMSE) of 1.6% for strain tracking using RFMLA on random wave pattern data.
  • Smart Shirt Accuracy: The integrated system estimated trunk flexion, lateral bending, and rotation angles with average errors of 4.26°, 3.53°, and 3.44°, respectively.
  • High Correlation: The system achieved a high correlation (R²) between estimated and actual angles, with values above 0.91 for all angles averaged across participants.
  • Movement Complexity: Accuracy remained high for simple planar movements (R² > 0.97) and showed good performance even for complex multiaxial movements (R² > 0.82), with the lowest accuracy in the random combination movement condition.

Limitations and Future Work

  • Speed Dependence: While generally accurate, the system showed a slight decrease in accuracy with faster movements, particularly for rotation angles.
  • Movement Complexity: The accuracy decreased with increasing movement complexity, with the random combination movement condition exhibiting the highest errors.
  • Future Work: The authors suggest that further optimization of sensor placement, signal preprocessing, and specific data analysis approaches are needed to achieve even more promising results in tracking complex three-dimensional movements. They also highlight the potential for this technology in long-term, unobtrusive data collection for applications like personalized interventions for lower back pain.

Summary updated

Abstract

Wearable electronics are recognized as a vital tool for gathering in situ kinematic information of human body movements. In this paper, we describe the production of a core–sheath fiber strain sensor from readily available materials in a one-step dip-coating process, and demonstrate the development of a smart sleeveless shirt for measuring the kinematic angles of the trunk relative to the pelvis in complicated three-dimensional movements. The sensor’s piezoresistive properties and characteristics were studied with respect to the type of core material used. Sensor performance was optimized by straining above the intended working region to increase the consistency and accuracy of the piezoresistive sensor. The accuracy of the sensor when tracking random movements was tested using a rigorous 4-h random wave pattern to mimic what would be required for satisfactory use in prototype devices. By processing the raw signal with a machine learning algorithm, we were able to track a strain of random wave patterns to a normalized root mean square error of 1.6%, highlighting the consistency and reproducible behavior of the relatively simple sensor. Then, we evaluated the performance of these sensors in a prototype motion capture shirt, in a study with 12 participants performing a set of eight different types of uniaxial and multiaxial movements. A machine learning random forest regressor model estimated the trunk flexion, lateral bending, and rotation angles with errors of 4.26°, 3.53°, and 3.44° respectively. These results demonstrate the feasibility of using smart textiles for capturing complicated movements and a solution for the real-time monitoring of daily activities.
BibTeX
@inproceedings{rezaei2019application,
  author = {Ahmad Rezaei and Tyler J. Cuthbert and Mohsen Gholami and Carlo Menon},
  title = {{Application-Based Production and Testing of a Core-Sheath Fiber Strain Sensor for Wearable Electronics: Feasibility Study of Using the Sensors in Measuring Tri-Axial Trunk Motion Angles}},
  booktitle = {Sensors},
  year = {2019},
  url = {https://openreview.net/forum?id=RzhoVYeaqw},
}
Figure