neutouchNANJING UNIVERSITY
NEUTOUCH LAB / NANJING UNIVERSITY

Teaching
machines
to feel.

Exploring tactile intelligence through skin-inspired sensing, neuromorphic electronics, and the seamless integration of matter and computation.

Explore our research ↗Our approach ↗
FIG. 01 — INTERACTIVE TACTILE RESPONSE FIELD
MOVE YOUR POINTER TO APPLY CONTACT
PHYSICS → SENSING → INTELLIGENCEINTERACTIVE RESEARCH VISUALIZATION / 001
AT THE INTERSECTION / 01

At NeuTouch Lab, we study how physical contact becomes meaningful information — connecting mechanics, emerging devices, electronic circuits, and intelligent systems to build the next generation of tactile technology.

01 / RESEARCH DIRECTIONS

Four scales.
One sense of touch.

Our research moves across scales, from material deformation and sensor physics to efficient hardware and intelligent perception.

01

Neuromorphic Circuits

Neuromorphic Circuits
↗
原创科研概念示意图：Event-driven neural encoding diagram

Event-driven circuits that encode and process touch close to the sensor, inspired by biological neural systems.

02

Emerging Neuromorphic Devices

Emerging Neuromorphic Devices
↗
原创科研概念示意图：Multilayer tactile synaptic device diagram

Synaptic transistors that integrate sensing, memory and learning within a single physical element.

03

Hardware–Software Co-Design

Hardware–Software Co-Design
↗
原创科研概念示意图：Sensor and on-device computation diagram

Joint design of sensors, circuits and learning algorithms for efficient tactile intelligence.

04

Mechanics of Skin-like Materials

Mechanics of Skin-like Materials
↗
原创科研概念示意图：Flexible skin finite-element deformation diagram

Modeling soft-material deformation under contact to guide sensor design and interpret tactile stimuli.

OUR VISION

What if machines could feel the world — not just see it?

02 / RESEARCH FRAMEWORK

From contact
to cognition.

From physical signals to meaningful perception — four connected stages in a tactile intelligence system.

A continuous conceptual flow from a robotic fingertip touching a flexible sensor, through tactile signal encoding and neuromorphic computing, to robotic object understanding and a feedback loop.
FIGURE 02 / THE TACTILE INTELLIGENCE PIPELINEConcept illustration · not experimental dataView full figure ↗
  1. 01
    CONTACT & SIGNAL

    Sensing

    Skin-like tactile sensors convert pressure, contact position and deformation into measurable electrical signals.

    Electronic skin · Tactile arrays
  2. 02
    EVENT REPRESENTATION

    Encoding

    Continuous tactile signals become temporal features and sparse, event-driven spikes for efficient processing.

    Temporal features · Spike trains
  3. 03
    NEUROMORPHIC PROCESSING

    Computing

    Devices, circuits and algorithms cooperate to extract tactile features close to the sensor.

    Neuromorphic circuits · Edge AI
  4. 04
    PERCEPTION & FEEDBACK

    Understanding

    Tactile features inform object and contact-state recognition, with feedback guiding robotic interaction.

    Object recognition · Closed loop

A unified conceptual workflow for explaining research directions; not a claim that all depicted systems have been experimentally demonstrated by the lab.

03 / SELECTED PUBLICATIONS

Research made
tangible.

Peer-reviewed work by Fengyuan Liu and collaborators. Selected visual evidence is reproduced from the original open-access research—not AI-generated artwork.

RESEARCH HIGHLIGHT / 001ORIGINAL EXPERIMENTAL FIGURES
Original Figure 1: electronic skin on a robotic hand, biological and artificial tactile neural pathways
科学机器人学 · 2022 · 原创研究

用于机器人感知和学习的基于印刷突触晶体管的电子皮肤

一种柔性计算电子皮肤结合了印刷突触晶体管、事件驱动传感和脉冲神经元电路，使机器人手能够通过局部学习获得触觉反应。

12 × 14印刷晶体管阵列
100%报告的器件良率
2022科学机器人
阅读研究论文 ↗
来自同一篇已发表的研究

器件与实验验证

从柔性器件、阵列一致性到机器人触觉学习，展示同一研究中的三组实验结果。

Figure 2g presentation with background edited to white; original available on click

柔性印刷器件

聚酰亚胺衬底上的印刷器件可随基底弯曲。研究进一步表征了机械弯曲前后的器件性能。

Original Figure 2d through f, spatial maps of transistor on-current, threshold voltage and off-current

器件阵列均匀性

三幅空间分布图分别展示导通电流、阈值电压和关断电流，用于评估印刷晶体管阵列的器件一致性。

Original Figure 5f: robotic arm photographs before and after associative learning, showing acquired pain reflex

学习触觉反应

关联学习前，机器人未表现出痛觉反射；学习后，相同演示中的机器人产生了相应的反射动作。

更多已核验研究

更多论文。

论文信息直接链接到来源记录，此处不再额外展示图版。

02 / REVIEWSCIENCE ROBOTICS · 2022

Neuro-inspired electronic skin for robots

Reviewing how neuromorphic hardware, printed electronics and distributed sensory computation can support robotic electronic skin.

Read open manuscript ↗DOI 10.1126/scirobotics.abl7344
03 / RESEARCHADV. FUNCTIONAL MATERIALS · 2025

Stochastic Nature of Large-Scale Contact Printed ZnO Nanowires Based Transistors

Characterizing ZnO nanowire transistor variability and using statistical models to predict performance across printed devices.

View open-access paper ↗DOI 10.1002/adfm.202412299
04 / FROM THE LAB

Notes &
milestones.

People, ideas and events shaping the development of NeuTouch Lab.

OCT 2026Welcome master student Kaibo Wang to the group.↗
OCT 2026Fengyuan Liu attended the Micro/Nano Optical Technology and Application Conference in Chengdu.↗
SEP 2026Welcome bachelor students Yizhuo Zhao and Jingqi Yin.↗
SEP 2026Welcome master students Mohan Cao and Dingkun Xia.↗
JUL 2026NeuTouch Lab is founded.↗
View original news archive ↗
05 / JOIN NEUTOUCH LAB

Curiosity
into discovery.

We welcome motivated researchers interested in circuits, mechanics, robotics, device physics, materials and cross-disciplinary tactile intelligence.

OPEN TO COLLABORATION

Build what touch can become.

Explore research opportunities and reach out with your research interests and experience.

Circuit DesignTactile SensingRoboticsMechanicsDevices
Contact the lab ↗