Nankai University Monolithic Synaptic Device Replicates Tactile Sensing and Neuromorphic Computing

Breakthrough in Integrated Tactile Sensing and Brain-Like Processing

Researchers at Nankai University have unveiled a groundbreaking monolithic synaptic device that merges tactile sensing with neuromorphic computing in a single, compact structure. This innovation, detailed in a recent paper published in Cyborg and Bionic Systems, promises to transform how robots perceive and respond to gentle human touch, paving the way for more intuitive human-robot interactions. 40 41

The device, known as a Pressure-Electronic-Gated (PEG) synaptic transistor, emulates the functions of C-tactile afferents—specialized nerve fibers in human skin that detect soft, emotional touches like stroking or patting. By integrating sensing and computation on-chip, it eliminates the need for bulky external processors, reducing latency and power consumption significantly. This is particularly relevant for applications in social robotics, where understanding nuanced touch can foster empathy and trust. 52

At the heart of this advancement is Wentao Xu's team from Nankai's College of Electronic Information and Optical Engineering. Their work builds on earlier efforts, including a May 2025 publication on SnO₂ nanowire-based devices, showcasing consistent progress in thin-film photoelectronic technologies. 41

Neuromorphic Computing: Mimicking the Brain's Efficiency

Neuromorphic computing refers to hardware designs that replicate the neural architecture and processing of the human brain, using artificial synapses and neurons to handle data in parallel with ultra-low power. Traditional von Neumann architectures separate memory and processing, leading to the 'von Neumann bottleneck' that hampers speed and efficiency for AI tasks. In contrast, neuromorphic systems like synaptic transistors process information locally at the point of sensing. 53

Synaptic devices, often transistor-based, emulate synaptic plasticity—short-term (e.g., paired-pulse facilitation) and long-term (e.g., retention over minutes)—through ion migration or charge trapping. Nankai's device advances this by making plasticity directly tunable by mechanical pressure, enabling in-sensor tactile computing without additional circuitry.

• Energy savings: Operates at -0.2 V, with currents from 39 nA to 25 μA.
• Stability: Endures 1,000 touch cycles and 2,000 seconds of operation.
• Scalability potential: Monolithic design suits array integration for 'electronic skin'. 40

This aligns with global trends where neuromorphic hardware is projected to grow rapidly, driven by AI demands in edge computing and robotics. 91

Decoding C-Tactile Afferents: The Science of Affective Touch

C-tactile (CT) afferents are unmyelinated C-fibers in hairy human skin that respond optimally to slow, gentle stroking at skin temperature (around 32°C), evoking pleasant, emotional sensations crucial for social bonding, grooming, and comfort. Unlike fast-conducting Aβ fibers for discriminative touch (texture, edges), CTs prioritize affective quality, firing maximally at velocities of 3-10 cm/s and pressures below 0.4 N. 63 64

The Nankai device replicates this with a chitosan hydrogel gate dielectric that modulates channel conductance via pressure-induced ionic migration in a poly(3-hexylthiophene) (P3HT) semiconductor layer. Step-by-step: (1) Gentle pressure (≥80 Pa) compresses the hydrogel, forming an electric double layer; (2) Protons migrate, gating the channel; (3) Excitatory postsynaptic currents (EPSCs) encode touch parameters like duration, frequency, and intensity; (4) Patterns classify emotions via simple algorithms. 52

Device Architecture and Performance Metrics

The PEG synaptic transistor features gold source/drain electrodes, P3HT channel, chitosan hydrogel dielectric, and a top mechanical pressure input. Electrochemical tests confirm proton conductivity and capacitance changes with pressure. Key metrics include:

• Threshold pressure: 80 Pa (human-like sensitivity).
• Spike-rate dependent plasticity: EPSC rises with touch frequency, mimicking habituation.
• Multimodal encoding: Distinguishes clapping (high freq., happy), patting (medium, calm), rocking (low, soothing). 40

In demos, connected to a microcomputer, it achieves real-time emotional classification, outputting robot responses like smiling LEDs for 'happy' pats. Compared to prior sensors needing >1 kPa or high voltages, this is a leap in efficiency.Read the full paper 42

Photo by Pawel Czerwinski on Unsplash

Nankai University's Leadership in Neuromorphic Research

Nankai University, a top-tier institution in Tianjin, China, has a storied history in optoelectronics and AI. The Institute of Photoelectronic Thin Film Devices and Technology, led by Prof. Wentao Xu, has pioneered flexible synapses for motion perception (2023) and visual neuromorphic systems. 72 This PEG device extends their portfolio, supported by National Key R&D Programs.

In China's vibrant higher education landscape, Nankai contributes to national AI strategies, fostering talent for research jobs in neuromorphic hardware. Students and postdocs here gain hands-on experience in cutting-edge labs, ideal for careers in AI and robotics.

China's Neuromorphic Ecosystem and Global Context

China leads in neuromorphic advances, with universities like Tsinghua and Fudan developing memristor arrays and 2D material synapses. The market for neuromorphic computing in China is booming, fueled by 'Made in China 2025' and investments in edge AI. 84 Nankai's work complements this, addressing tactile gaps in vision-dominant systems.

Globally, IBM's TrueNorth and Intel's Loihi set benchmarks, but organic devices like Nankai's excel in flexibility for wearables. Challenges include uniformity in arrays and endurance, yet progress is swift. 82

Real-World Applications in Robotics and Beyond

Imagine robots in elderly care responding affectionately to hugs or therapy bots for autism aiding emotional regulation via calibrated touch. The device enables 'affective computing'—machines that sense emotions—crucial for humanoid robots like Tesla's Optimus. 101

• Social assistive robotics: Enhances empathy in interactions.
• Prosthetics: Provides natural feedback for amputees.
• VR/AR: Haptic interfaces for immersive experiences.

For aspiring engineers, explore academic CV tips to join such projects via China university jobs.

Overcoming Challenges: Scalability and Energy Hurdles

While promising, neuromorphic devices face scalability issues—fabricating uniform large arrays—and variability in organic materials. Energy efficiency is stellar (pJ per synapse), but integration with silicon remains key. 92 Nankai's hydrogel-P3HT stack shows retention >10^3 s, but future hybrid 2D materials could boost it.

Solutions include AI-optimized training and standardized benchmarks, positioning China universities at the forefront.

Photo by Logan Voss on Unsplash

Future Outlook: Toward Intelligent Electronic Skin

Prof. Xu envisions flexible arrays for full-body robot skin, enabling spatial touch mapping. Coupled with multimodal sensing (vision + touch), this could yield artificial nervous systems. 40 In higher education, it spurs demand for faculty positions in AI materials science.

As China invests billions in AI, Nankai's innovation underscores universities' role in bridging academia and industry. Researchers eyeing postdocs should check postdoc opportunities.

Career Implications in China's AI Boom

This breakthrough highlights booming opportunities in China's neuromorphic sector. Graduates from Nankai and peers can pursue roles in robotics firms like UBTech or state labs. Platforms like AcademicJobs.com higher-ed-jobs list relevant openings, including research assistant jobs.

For career advice, visit higher-ed-career-advice or rate-my-professor for insights on mentors like Prof. Xu.