Recent Advances in Flexible Tactile Sensors for Intelligent Systems

Yiyao Peng, Ning Yang, Qian Xu, Yang Dai, Zhiqiang Wang2021 · DOI

TL;DR

This review paper provides a comprehensive overview of recent advancements in flexible tactile sensors, focusing on their transduction mechanisms, performance enhancements (high resolution, sensitivity, self-powering, visualization), and diverse applications in intelligent systems like robotics, wearables, and human-computer interfaces. It highlights the growing importance of these sensors in bridging the gap between artificial intelligence and physical interaction.

What Problem Does This Paper Address?

Traditional pressure sensors are often rigid, limiting their use in flexible and wearable applications. The increasing demand for intelligent systems that can interact with the physical world—such as advanced robots, sophisticated prosthetics, and intuitive human-computer interfaces—requires sensors that can mimic the nuanced touch capabilities of human skin. This paper addresses the need for flexible tactile sensors that are not only highly performant (sensitive, high-resolution) but also multifunctional (self-powered, visual feedback) to enable these next-generation intelligent systems.

Methods

The paper categorizes flexible tactile sensors based on their primary transduction mechanisms:

  • Piezoresistive: Resistance changes with applied pressure. This is a widely researched area using materials like carbon nanotubes, graphene, and various composite fillers embedded in flexible substrates.
  • Capacitive: Capacitance changes due to variations in distance or overlap area between electrodes. Techniques include using micro/nanostructured dielectric layers and flexible electrode networks.
  • Piezoelectric: Generates an electrical charge when subjected to mechanical stress. This often involves piezoelectric polymers (like PVDF) or inorganic materials (like ZnO, GaN) in various nanostructures.
  • Triboelectric: Generates electricity through the triboelectric effect (frictional electrification) and electrostatic induction. Triboelectric nanogenerators (TENGs) are a key technology here, enabling self-powered sensors.

The review also discusses advancements in sensor performance, including:

  • High Resolution: Achieved through piezoelectric effects in nanostructures, enabling resolutions finer than human skin.
  • High Sensitivity: Detecting very low pressures (e.g., 3 Pa) using novel materials and structures like microstructured PDMS.
  • Self-Powered: Utilizing triboelectric or piezoelectric effects to generate power from ambient mechanical energy, eliminating the need for external power sources.
  • Visualization: Integrating sensors with visual output mechanisms (e.g., organic electrochromic devices, LED arrays) to provide direct visual feedback of tactile information.

Datasets

Not discussed in the paper. The paper is a review of existing research and does not present new experimental data or datasets.

Key Insights and Ideas

  • Mimicking Human Skin: The ultimate goal is to create "electronic skin" that replicates the sensitivity, resolution, and multi-modal sensing capabilities of human skin.
  • Multifunctionality is Key: Beyond just sensing pressure, future tactile sensors need to be self-powered, provide visual feedback, be biocompatible, and even self-healing to be truly useful in intelligent systems.
  • Nanomaterials Drive Performance: The use of nanomaterials like carbon nanotubes, graphene, nanowires, and nanoplatelets is crucial for achieving high sensitivity and resolution in flexible tactile sensors.
  • Triboelectricity for Self-Powering: TENGs are a significant breakthrough, enabling self-powered tactile sensors that are essential for long-lasting, low-maintenance intelligent devices.
  • Integration is the Future: The trend is towards integrating multiple sensing modalities (pressure, temperature, strain, etc.) into single, flexible sensor networks for comprehensive environmental perception.

Key Results

  • High Resolution Achieved: Piezoelectric-based sensors have demonstrated resolutions up to 12,700 dpi, exceeding human skin capabilities.
  • Ultra-High Sensitivity Demonstrated: Sensors capable of detecting pressures as low as 3 Pa have been developed.
  • Self-Powered Sensors are Viable: Triboelectric and piezoelectric sensors can generate sufficient power for their own operation and even for powering small electronic devices.
  • Visual Feedback Integrated: Chameleon-inspired systems and LED arrays can translate tactile input into visible color changes or light patterns.
  • Diverse Applications: Flexible tactile sensors are finding applications in wearable electronics for health monitoring, intelligent robots for object manipulation and navigation, human-computer interfaces for enhanced interaction, and implantable electronics for medical diagnostics and therapies.

Limitations and Future Work

  • Performance Degradation: Sensors can degrade over time due to repeated deformation, necessitating research into more robust materials and advanced packaging.
  • Crosstalk Decoupling: Simultaneously acquiring multiple signals (e.g., pressure, temperature, strain) without interference remains a challenge, with transistor structures being a promising solution.
  • Material and Component Matching: Ensuring that mechanical, thermal, and electrical properties of all components within an integrated sensing system are compatible is crucial.
  • Future Directions: The paper points towards continued development in flexibility, miniaturization, intelligence, and human-like functionality. Advanced materials, packaging, and integration techniques will be key. The integration of tactile sensors into artificial intelligence systems is still in its early stages, presenting a significant research direction for enabling more sophisticated human-robot interaction.

Summary updated

BibTeX
@misc{peng2021recent,
  author = {Yiyao Peng and Ning Yang and Qian Xu and Yang Dai and Zhiqiang Wang},
  title = {{Recent Advances in Flexible Tactile Sensors for Intelligent Systems}},
  year = {2021},
  doi = {10.3390/s21165392},
}
Figure