Project Tasca : Enabling Touch and Contextual Interactions with a Pocket-based Textile Sensor

Title of the Paper

Project Tasca: Enabling Touch and Contextual Interactions with a Pocket-based Textile Sensor

Paper Information

• Subject Area: Wearable technology, human-computer interaction, smart fabric technology
• Keywords: Capacitive sensing, inductive sensing, resistive sensing, NFC, interactive fabrics, gesture recognition, object recognition, wearables, interactive pockets

Research Background and Issues

• Identified Problems and Challenges:

  • The interaction design of wearable devices limits user input methods, such as through smartphones or smartwatches.
  • Existing interactive pocket designs rely on rigid devices (e.g., cameras, touch sensors), which are unsuitable for daily use and fail to achieve fabric flexibility and portability.

• Importance of the Problem:

  • Research on smart fabrics opens new possibilities for object interaction in daily life and work, especially enhancing users' ability for eyes-free, prolonged interactions.
  • Recognizing objects within pockets provides new scenarios for activity tracking, environmental adjustment, and information integration between devices.

• Research Motivation and Related Work:

  • Previous studies have demonstrated the potential of interactive fabrics but have not yet combined multiple sensing technologies to provide a universal solution.
  • This project aims to improve flexible smart fabric design to support various interactions, ranging from touch gestures to object recognition and contextual awareness.

Solution

• Proposed Method:

  • Develop an interactive pocket based on a multi-layer textile sensor, integrating four sensing technologies: inductive sensing, capacitive sensing, resistive sensing, and NFC.
  • Design hardware integration to achieve lightweight sensor deployment within the fabric.

• Innovations:

  • For the first time, multiple sensing technologies are combined to integrate object recognition, touch gestures, pressure sensing, and NFC identification into a single textile sensing system.
  • A multi-layer fabric design reduces thickness and optimizes sensor arrangement to enhance sensing accuracy and expand application scenarios.

• Implementation Steps and Key Technologies:

  • Inductive Sensing: Use spiral coils to detect metal objects by sensing changes in the electromagnetic field.
  • Capacitive Sensing: Detect the "capacitive footprint" of non-metallic objects and use it for touch gesture recognition.
  • Resistive Sensing: Employ pressure-sensitive materials to detect pressure, enabling recognition of object thickness and gesture intensity.
  • NFC Sensing: Identify objects with attached tags through electromagnetic field detection and communication.
  • Utilize a layered structure in sensor manufacturing to ensure sensing accuracy and device flexibility.

Research Outcomes

• Specific Results:

  • The pocket sensor achieved a recognition accuracy of 92.3% for object identification and 96.4% for gesture recognition, reliably detecting NFC tags with a success rate of 100%.
  • Enables input through gestures and various pressure levels while providing contextual interactions via object recognition.

• Advantages over Existing Solutions:

  • Compared to existing pocket designs using rigid devices, this system is more portable, flexible, and suitable for daily use.
  • The combination of multiple sensors improves data reliability and adapts to individual user differences.

• Experimental and Evaluation Results:

  • Object Recognition Experiment: Tested 11 objects, including empty and filled containers, achieving cross-user accuracy of 81.3%.
  • Gesture Recognition Experiment: Tested 8 common touch gestures with an accuracy of 96.1%. Pressure gestures effectively distinguished between low and high pressure.
  • NFC Sensing Experiment: Tags were accurately identified within 0-6 mm for distance and position, though remote detection was limited.

• Limitations and Future Directions:

  • Hardware Limitations: Sensor arrangement and resolution restrict recognition of certain objects, requiring hardware improvements such as multiplexers.
  • Impact of Body Movements: Sensor readings may be noisy during user motion, necessitating more accurate models to process dynamic input.
  • Application Scenario Expansion: Explore other clothing scenarios beyond pockets and non-contact detection methods.
  • Fabric Flexibility Optimization: Enhance fabric softness to improve the practical wearability experience for users.

Conclusion

This study demonstrates the novel interaction capabilities of multi-layer fabric through multi-sensor integration, supporting touch, pressure, and object recognition with high accuracy and contextual adaptability across various application scenarios. Future work will focus on hardware optimization and exploring new application areas to advance the practicality and expansion of smart fabric technology.