aiqbench analysis
This research presents a hybrid tactile sensor that combines electrical impedance tomography with pneumatic pressure layers to significantly improve force estimation accuracy in robotic skins. By leveraging the complementary strengths of both sensing modalities, the system reduces root-mean-square error by up to 60% while avoiding the need for heavy machine-learning pipelines or large datasets. This approach offers a practical, scalable solution for building robust, large-area contact sensors essential for advanced robotics applications.
Pneumatic-Tomographic Tactile Skin for Multicontact Localization and Force Estimation
arXiv:2503.13036v3 Announce Type: replace Abstract: Tactile skins based on electrical impedance tomography (EIT) enable large-area contact localization with few electrodes, but suffer from nonuniform sensitivity that limits force estimation accuracy. This work introduces a dual-channel tactile skin that integrates an EIT layer with a pneumatic pressure layer and a calibration framework that leverages their complementary strengths. The EIT layer provides robust multicontact localization, while the pneumatic pressure layer supplies a stable scalar measurement that serves as contact force estimation. A location-aware correction method is introduced, learning smooth spatial gain and offset fields from a single-session calibration, enabling spatially consistent multicontact force estimation. With location-aware correction, the proposed system achieves a single-contact force estimation root-mean-square error (RMSE) of 0.59 N across 10-25-mm indenters, representing a 35%-60% reduction over EIT-only approaches (1.45-1.48 N). In multicontact experiments, the per-contact RMSE is reduced by 39.6% compared to the uncorrected pneumatic baseline. The proposed system achieves accurate force estimation across diverse contact configurations, generalizes to varying indenter sizes, and preserves EIT's inherent advantages in multicontact localization. By letting the pneumatic pressure layer handle the force estimation and using the EIT layer to determine where each contact occurs, the method avoids the need for large datasets, complicated calibration setups, and heavy machine-learning pipelines often required by previous EIT-only approaches. This dual-channel design provides a practical, scalable, and easy-to-calibrate solution for building large-area robotic skins.
Abstract:Tactile skins based on electrical impedance tomography (EIT) enable large-area contact localization with few electrodes, but suffer from nonuniform sensitivity that limits force estimation accuracy. This work introduces a dual-channel tactile skin that integrates an EIT layer with a pneumatic pressure layer and a calibration framework that leverages their complementary strengths. The EIT layer provides robust multicontact localization, while the pneumatic pressure layer supplies a stable scalar measurement that serves as contact force estimation. A location-aware correction method is introduced, learning smooth spatial gain and offset fields from a single-session calibration, enabling spatially consistent multicontact force estimation. With location-aware correction, the proposed system achieves a single-contact force estimation root-mean-square error (RMSE) of 0.59 N across 10-25-mm indenters, representing a 35%-60% reduction over EIT-only approaches (1.45-1.48 N). In multicontact experiments, the per-contact RMSE is reduced by 39.6% compared to the uncorrected pneumatic baseline. The proposed system achieves accurate force estimation across diverse contact configurations, generalizes to varying indenter sizes, and preserves EIT's inherent advantages in multicontact localization. By letting the pneumatic pressure layer handle the force estimation and using the EIT layer to determine where each contact occurs, the method avoids the need for large datasets, complicated calibration setups, and heavy machine-learning pipelines often required by previous EIT-only approaches. This dual-channel design provides a practical, scalable, and easy-to-calibrate solution for building large-area robotic skins.
Robotics (cs.RO); Signal Processing (eess.SP)
I.2.9; I.4.5; I.4.8
arXiv:2503.13036 [cs.RO]
(or arXiv:2503.13036v3 [cs.RO] for this version)
Chen, H., Kubik, J., Himmel, B., Hoffmann, M. and Lee, H. (2026), 'Pneumatic-Tomographic Tactile Skin for Multicontact Localization and Force Estimation', IEEE Sensors 26(16), 23655 - 23665
https://doi.org/10.1109/JSEN.2026.3704916 DOI(s) linking to related resources
From: Matej Hoffmann Ph.D. [view email] [v1] Mon, 17 Mar 2025 10:41:13 UTC (1,451 KB) [v2] Thu, 18 Dec 2025 18:54:47 UTC (1,099 KB) [v3] Thu, 27 Aug 2026 08:55:40 UTC (1,312 KB)
Full text under the publisher's open licence · 305 words
Related coverage
A Wearable Pneumatic Device for Continuous, Closed-Loop, Bidirectional Tactile Interaction
arXiv · Sep 2, 2026
Tactile-JEPA: Topology-Aware Self-Supervised Representation Learning for Distributed Tactile Sensors
arXiv · Sep 22, 2026
External Photoreflective Tactile Sensing Based on Surface Deformation Measurement
arXiv · Oct 2, 2026
Tactile Perception through Fluid-Solid Interaction
arXiv · Oct 6, 2026