Intelligent Recognition Using Ultralight Multifunctional Nano-Layered Carbon Aerogel Sensors with Human-Like Tactile Perception

Huiqi Zhao, Yizheng Zhang, Lei Han, Weiqi Qian, Jiabin Wang, Heting Wu, Jingchen Li, Yuan Dai, Zhengyou Zhang, Chris Bowen, Ya YangNano-Micro 2023 · DOI

Abstract

Humans can perceive our complex world through multi-sensory fusion. Under limited visual conditions, people can sense a variety of tactile signals to identify objects accurately and rapidly. However, replicating this unique capability in robots remains a significant challenge. Here, we present a new form of ultralight multifunctional tactile nano-layered carbon aerogel sensor that provides pressure, temperature, material recognition and 3D location capabilities, which is combined with multimodal supervised learning algorithms for object recognition. The sensor exhibits human-like pressure (0.04-100 kPa) and temperature (21.5-66.2 °C) detection, millisecond response times (11 ms), a pressure sensitivity of 92.22 kPa<sup>-1</sup> and triboelectric durability of over 6000 cycles. The devised algorithm has universality and can accommodate a range of application scenarios. The tactile system can identify common foods in a kitchen scene with 94.63% accuracy and explore the topographic and geomorphic features of a Mars scene with 100% accuracy. This sensing approach empowers robots with versatile tactile perception to advance future society toward heightened sensing, recognition and intelligence.

Introduction

While robotic vision has advanced significantly, it remains inherently brittle in environments with poor visibility, such as smoke-filled rooms, low light, or Martian dust storms. Furthermore, cameras alone cannot distinguish between objects with similar appearances but different physical properties—such as a plastic fruit versus a real one. To bridge this gap, researchers have developed an ultralight, multifunctional tactile aerogel sensor (MTAS) that mimics the multi-sensory fusion of human skin. By integrating pressure, temperature, and material recognition into a single unit, this technology enables robots to "feel" their surroundings with high precision, even in the absence of visual data.

Sensor Design and Methodology

The core innovation lies in the material composition and internal architecture of the sensor. The MTAS is constructed from a reduced graphene oxide (rGO) and cellulose nanocrystal (CNC) carbon aerogel. Using a directional freezing and annealing technique, the team achieved a unique "wavy-layered" internal microstructure. This morphology acts as a microscopic spring, allowing the sensor to detect pressures as low as 30 Pa (roughly the weight of a small pill) and withstand significant strain up to 90%.

Schematic of the MTAS architecture showing the layered structure (PE film, Cu film, Aerogel, Ecoflex) and its ability to sense pressure, temperature, and material composition.

The sensor operates through three distinct physical mechanisms simultaneously:

  1. Piezoresistive: Changes in resistance as the wavy layers compress allow for precise pressure detection.
  2. Thermoelectric: The sensor generates a voltage from temperature gradients, enabling thermal sensing.
  3. Triboelectric: Contact electrification between the sensor's outer polyethylene (PE) film and the object being touched allows for material identification.

To process these overlapping signals, a multimodal supervised learning framework was developed. It utilizes parallel 1D Convolutional Neural Networks (CNNs) to extract features from time-variant signals, which are then fused by a Multi-Layer Perceptron (MLP) for final classification.

Experimental Evaluation and Results

The system was evaluated using three primary datasets: a "Kitchen Dataset" featuring 18 types of food (e.g., banana, meat, shrimp), a "Mars Landform Dataset" featuring four distinct terrains (lake, sand, soil, and stone), and a material characterization benchmark of 11 standard materials.

Application scenarios for the tactile system, including food recognition in a kitchen and terrain analysis in a simulated Mars environment.

The MTAS demonstrated exceptional performance across all benchmarks:

  • Recognition Accuracy: The system achieved 94.63% accuracy in identifying kitchen ingredients and a perfect 100% accuracy in distinguishing Martian landforms.
  • Physical Performance: The sensor is incredibly light (64.5 mg) and features a rapid response time of 11 ms with a high pressure sensitivity of 92.22 kPa⁻¹.
  • Durability: It maintained stable performance over 6,000 cycles of triboelectric testing and 700 cycles of high-strain compression.
  • Efficiency: Both temperature and material recognition modes are self-powered, which could significantly reduce the power budget for complex robotic skins.

Conclusion and Future Outlook

The MTAS represents a significant step toward human-like tactile perception in robotics. By combining advanced material science with multimodal deep learning, the researchers have created a sensor that is not only lightweight and sensitive but also capable of complex object recognition without visual aid.

Future work will focus on increasing sensor density to provide better spatial resolution across robotic surfaces. Additionally, the researchers suggest that the triboelectric properties of the MTAS could be used for active surface maintenance, such as repelling dust from lenses or solar panels—a critical feature for long-term space exploration. Further testing is also required to ensure the sensor's survival in the extreme cryogenic temperatures and radiation of actual planetary surfaces.

Summary updated

Abstract

Humans can perceive our complex world through multi-sensory fusion. Under limited visual conditions, people can sense a variety of tactile signals to identify objects accurately and rapidly. However, replicating this unique capability in robots remains a significant challenge. Here, we present a new form of ultralight multifunctional tactile nano-layered carbon aerogel sensor that provides pressure, temperature, material recognition and 3D location capabilities, which is combined with multimodal supervised learning algorithms for object recognition. The sensor exhibits human-like pressure (0.04-100 kPa) and temperature (21.5-66.2 °C) detection, millisecond response times (11 ms), a pressure sensitivity of 92.22 kPa<sup>-1</sup> and triboelectric durability of over 6000 cycles. The devised algorithm has universality and can accommodate a range of application scenarios. The tactile system can identify common foods in a kitchen scene with 94.63% accuracy and explore the topographic and geomorphic features of a Mars scene with 100% accuracy. This sensing approach empowers robots with versatile tactile perception to advance future society toward heightened sensing, recognition and intelligence.
BibTeX
@inproceedings{zhao2023intelligent,
  author = {Huiqi Zhao and Yizheng Zhang and Lei Han and Weiqi Qian and Jiabin Wang and Heting Wu and Jingchen Li and Yuan Dai and Zhengyou Zhang and Chris Bowen and Ya Yang},
  title = {{Intelligent Recognition Using Ultralight Multifunctional Nano-Layered Carbon Aerogel Sensors with Human-Like Tactile Perception}},
  booktitle = {Nano-Micro},
  year = {2023},
  doi = {10.1007/s40820-023-01216-0},
}
Figure