Neuromorphic high-density tactile sensor readout for electronic skins (e-skins)

Mark Alea , Georges Gielen Biomedical circuits and sensor interfaces

To mimic the rich tactile information captured by human hands and fingertips during object manipulation in electronic skin (e-skin) applications such as for prosthetics and robotics, it is necessary to achieve tactile sensing and processing with at least the same energy efficiency and taxel density. The research aims for this goal by integrating a dense tactile sensor array on a custom readout chip, allowing for dense and power-efficient readout with local processing capabilities. The sensor-to-readout chip integration supports dense per-taxel connections, enabling true per-taxel sparse-sampling conversion (i.e., spikes) for further local signal processing. Two versions of the taxel readout design have been developed.
The first design is an e-skin taxel readout chip in 0.18μm CMOS technology has been realized that achieves the highest reported spatial resolution of 200μm, comparable to human fingertips. PVDF-based piezoelectric sensors are integrated on the chip, enabling a per-taxel signal conditioning frontend and spiking readout combined with local embedded neuromorphic first-order processing through complex receptive fields (CRFs). Experimental results show that Spiking Neural Network (SNN)-based offline classification of the chip’s spatiotemporal spiking output for input tactile stimuli such as texture and flutter frequency achieves excellent accuracies up to 97.1% and 99.2%, respectively.
The second design is an e-skin taxel readout chip in a-IGZO thin-film transistor (TFT) technology. It has lower spatial resolution, but has been designed to enable slip detection during grasping.

Get in touch
Mark Alea
Phd student
Georges Gielen
Academic staff
 (a) Equipping a robotic hand with a combination of large-area e-skin in the palm and high-spatial-resolution e-skin on the fingertips allows the classification of b) macro and micro tactile features. (c) The receptor density and complex receptive fields in human fingertips are mimicked in the presented d) INTUITIVE e-skin sensing chip.
(a) Equipping a robotic hand with a combination of large-area e-skin in the palm and high-spatial-resolution e-skin on the fingertips allows the classification of b) macro and micro tactile features. (c) The receptor density and complex receptive fields in human fingertips are mimicked in the presented d) INTUITIVE e-skin sensing chip.

Publications about this research topic

M. D. Alea et al., "A Fingertip-Mimicking 12×16 200μm-Resolution e-skin Taxel Readout Chip with per-Taxel Spiking Readout and Embedded Receptive Field Processing," in IEEE Transactions on Biomedical Circuits and Systems, December 2024, doi: 10.1109/TBCAS.2024.3387545.

M. D. Alea et al., "A Fingertip-Mimicking 12×16 200μm-Resolution e-skin Taxel Readout Chip with per-Taxel Spiking Readout and Embedded Receptive Field Processing," in proceedings IEEE Symposium on VLSI Technology and Circuits, 2023, doi: 10.23919/VLSITechnologyandCir57934.2023.10185346.

M. D. Alea, A. Safa, J. V. Assche and G. G. E. Gielen, "Power-Efficient and Accurate Texture Sensing Using Spiking Readouts for High-Density e-Skins," in proceedings IEEE Biomedical Circuits and Systems Conference (BioCAS), 2022, doi: 10.1109/BioCAS54905.2022.9948546.

A. Safa, J. Van Assche, M. D. Alea, F. Catthoor and G. G. E. Gielen, "Neuromorphic Near-Sensor Computing: From Event-Based Sensing to Edge Learning," in IEEE Micro, Nov.-Dec. 2022, doi: 10.1109/MM.2022.3195634.

Other research topics in Biomedical circuits and sensor interfaces

Scalable large array nanopore readouts for proteomics and next-generation sequencing
Analog and power management circuits, Hardware-efficient AI and ML, Biomedical circuits and sensor interfaces
Sander Crols | Filip Tavernier and Marian Verhelst
Implantable neural readout frontend design for closed-loop neuromodulation systems
Mixed-signal circuits and data converters, Biomedical circuits and sensor interfaces
Marco Carlino | Georges Gielen
Ultrasound wave based body area networks
Analog and power management circuits, Ultra-low power digital SoCs and memories, Biomedical circuits and sensor interfaces
Wim Dehaene and Marian Verhelst
High-resolution neurostimulator ICs for the Visual Cortex
Biomedical circuits and sensor interfaces
Maxime Feyerick | Wim Dehaene

Want to work with us?

Get in touch or discover the way we can collaborate.