Smash

Objectives

Integrating various sensors and computing elements into clothing opens up a wide range of possible applications from new forms of consumer electronics to new innovative user interfaces. One important and promising aspect is medical applications using smart clothing. Real time sensor data can be used for health monitoring in an unobtrusive manner.

The motivation for this project was to develop a new versatile sensor platform, that was to be integrated into a common loose fitting long sleeve shirt. If the clothing is equipped with motion sensors, it can provide feedback about the wearer’s movements or postures. Such information is helpful in rehabilitation or sport applications, where it is important that certain movements are executed correctly. For example rehabilitation exercises need to be performed in clearly a defined motion sequence, with the correct speed and a defined amount of repetitions.

Achievements

• A novel aspect is the straight focus on the wearability by using an everyday piece of clothing. The goal was not to develop a specialized textile that can only be used for one application, but rather to produce a generic system that can be customized for different applications, like pattern recognition or context recognition.

• A methodology for textile integration by using different silicone compounds has been developed and described. During the experimental phase the resulting shirt has proven to be comfortable and flexible, which allows unobtrusive data collection during various activities.

• The prototype shirt was used to determine system characteristics like its electrical properties and achievable data transfer rates. Experiments were conducted with several users posing in defined postures and limb articulations. The recorded data were used to characterize the accelerometer Terminals, as well as for input of classification algorithms, demonstrating the ability of the shirt to act as a pattern recognition (in this case posture recognition) system.

Publications

• SMASH: A Distributed Sensing and Processing Garment for the Classification of Upper Body Postures; H. Harms, O. Amft, D. Roggen, G. Tröster; Submitted to the third International Conference on Body Area Networks (BodyNets08), Tempe AZ, March 13-17, 2008.