Estimating Smart Garment Optimal Fitting during Activities of Daily Living

Symbolbild zum Artikel. Der Link öffnet das Bild in einer großen Anzeige.

Nowadays the long-term monitoring of health human parameters can be achieved by using wearable devices. For most health monitoring applications, the position of the sensor on the body plays a major role and, if the sensor is not placed correctly, the wearable system performance can be seriously affected.

Among the wearable devices, smart clothes have been extensively investigated for their comfort and ease of use. In smart clothes, the sensitivity and performance of the embedded sensors depends not only on the relative body-sensor position, but also on the fitting. A loose skin-sensor interface can compromise the device functionality, while a bad
fitting of the body shape can affect its usability. It is therefore important to assess how and to what extent the fitting influences and impacts the garment sensing capabilities.

Over the last decades, different problem modelling approaches were proposed to support developers in the garment design. However, all the analysis were conducted only in human static positions, without taking into account sensor errors introduced during motion.

In this project, we aim to implement a framework to automatically design a personalised smart garment, including embedded sensors, and evaluate performance changes caused by human motion from Activities of Daily Living (ADL) for different sensing modalities.

Innovation

We developed a framework to automatically extract body landmarks and related measurements from 3D body scans and generate a personalised smart garment based on the estimated body measurements. We trained and tested the algorithm on 3000 synthetic 3D body models and estimated body landmarks required for T-Shirt design. We validated the framework the framework in automated tailoring of an electrocardiogram (ECG)-monitoring shirt based on the predicted landmarks. The ECG shirt could fit all evaluated body shapes with an average electrode-skin distance of 0.61 cm.

We designed a simulation method to evaluate the performance of garment-embedded contact sensors while performing common ADL. We generated 100 3D human body models with varying body shapes and virtually dressed them with three differently fitted smart ECG-monitoring T-Shirts. We then analysed the sensor-body distance and sensor displacement while performing common ADLs. We showed that the performance changes as a function of body shape, garment fit and movement type (ADL).

In further work we plan to extend the existing method to include fabric properties (e.g. cotton, stretchable/functional materials) to further investigate the impact on fit and sensor performance. Additionally, simulation of different garment cuts will further enhance the capabilities and scope of application of the method. Furthermore, other
sensor modalities (e.g. orientation and strain sensors) and corresponding evaluation routines will be included, to accommodate a greater variety of embedded electronics in garments.

Videos

Award

The paper “Simulation of Garment-Embedded Contact Sensor Performance under Motion Dynamics” presented by Annalisa Baronetto at Ubicomp/ISWC 2021 received the ‘Honorable Mention’ award.

Work with us

We offer interesting challenges for students interested in working with smart textiles, software engineering and CAD design.

Publications

Annalisa Baronetto, Dominik Wassermann, Oliver Amft, "Deep 3D Body Landmarks Estimation for Smart Garments Design", 2021 IEEE 17th International Conference on Wearable and Implantable Body Sensor Networks, 27-30 July 2021.
Annalisa Baronetto, Lena Uhlenberg, Dominik Wassermann, Oliver Amft, "Simulation of Garment-Embedded Contact Sensor Performance under Motion Dynamics", ACM International Symposium on Wearable Computers, 21-26 September 2021.

Contact

Annalisa Baronetto

  • Job title: Researcher
  • Organization: Department of Medical Informatics, Biometry and Epidemiology
  • Working group: Chair of Digital Health
  • Phone number: +49 9131 85 23608
  • Email: annalisa.baronetto@fau.de

Lena Uhlenberg

  • Job title: Researcher
  • Organization: Department of Medical Informatics, Biometry and Epidemiology
  • Working group: Chair of Digital Health
  • Phone number: +49 9131 85-23605
  • Email: lena.uhlenberg@fau.de

Friedrich-Alexander-Universität Erlangen-Nürnberg