Engineers have developed a dual-modal flexible textile sensor designed to capture both tensile strain and pressure during athletic movements, specifically targeting basketball-holding mechanics. Built upon a knitted Kevlar fabric substrate, the wearable technology applies a polyimide solution via dip coating followed by laser direct writing. This process induces in-situ graphitization of the polyimide to construct a three-dimensional porous laser-induced graphene conductive network directly on the fabric surface.
Fabrication and material characterization of the textile sensor
Researchers evaluated the structural and material properties of the conductive layer using scanning electron microscopy, thermogravimetric analysis, and Raman spectroscopy. These evaluations confirmed the successful formation of the porous graphene network on the durable Kevlar substrate. Tensile and pressure tests demonstrated that the sensor maintains stable performance across diverse mechanical deformations. In tensile sensing mode, the gauge factor achieved a maximum of 48.5, accompanied by a response time of 0.25 seconds and a recovery time of 0.22 seconds. In pressure sensing mode, the sensitivity reached a maximum of 1.22 per kilopascal, with a response time of 0.03 seconds and a recovery time of 0.05 seconds. Both operational modes maintained stable electrical responses throughout cyclic testing lasting thousands of seconds.
Smart glove neural network achieves high posture recognition accuracy
To test real-world applications in athletic monitoring, the research team integrated a sensor array into a wearable smart glove. The glove collected multichannel resistance signals as test subjects executed four typical basketball-holding postures. These signals were processed using a one-dimensional convolutional neural network to classify the hand positions. The deep learning model achieved a posture recognition accuracy of 94 percent across the tested grip configurations. This performance highlights the potential for combining flexible textile sensors with machine learning algorithms for human-machine interaction and smart sports analytics.
Performance metrics of the Kevlar-graphene sensor
- Tensile Gauge Factor: Reached a maximum of 48.5 for strain monitoring.
- Tensile Response/Recovery: 0.25 seconds response time and 0.22 seconds recovery time.
- Pressure Sensitivity: Reached a maximum of 1.22 per kilopascal.
- Pressure Response/Recovery: 0.03 seconds response time and 0.05 seconds recovery time.
- Posture Recognition Accuracy: 94 percent using a one-dimensional convolutional neural network.
Frequently asked questions about the smart sports sensor study
What materials make up the flexible substrate of the sensor?
The sensor uses a knitted Kevlar fabric as its flexible substrate, combined with a laser-induced graphene conductive network formed via polyimide solution dip coating.
How does the smart glove capture basketball-holding postures?
A sensor array embedded within the glove collects multichannel resistance signals as the wearer grips a basketball, which are then analyzed by a one-dimensional convolutional neural network.
What accuracy did the neural network achieve during testing?
The one-dimensional convolutional neural network achieved an overall posture recognition accuracy of 94 percent when evaluating four typical basketball-holding postures.
Worth a look
