flexible sensor integration

该研究通过在可拉伸液态金属互连中调制数字信号幅度，实现同时进行数字通信和变形感知，无需额外的传感元件。

This study addresses the sensing unreliability of flexible piezoresistive sensors caused by nonlinearity, hysteresis, and over-range conditions. We propose a reliability framework integrating physical constraints with uncertainty quantification to enable adaptive, trustworthy strain perception without requiring fault samples. By constructing a physics-guided probabilistic inverse model alongside a tri-state risk monitor for strain limits, the system achieves robust performance under complex operating conditions. Experimental results demonstrate a model fitting coefficient of 0.95, an RMSE as low as 0.15%, a 100% anomaly detection rate, and 96% empirical coverage. This work effectively mitigates sensing failures in demanding environments, significantly enhancing the robustness and trustworthiness of flexible sensing systems.

To address the challenge of real-time shape sensing in continuum robots, this paper proposes a novel strategy for fabricating epidermal flexible strain sensors via direct ink writing (DIW). For the first time, a high-viscosity Ga–In liquid metal ink is employed in DIW-based additive manufacturing to fabricate resistive microscale sensing traces, enabling millimeter-scale integration and minute-scale rapid fabrication. The resulting sensor exhibits near-zero drift, high linearity (R² ≈ 0.99), a stable gauge factor (GF ≈ 1), a broad strain-sensing range, and excellent repeatability. This approach overcomes key limitations of conventional sensor integration—such as bulkiness, poor conformability, and complex assembly—thereby delivering a lightweight, high-fidelity, and easily deployable deformation feedback solution tailored for continuum robots.

Existing resistive tactile gloves rely on manual assembly or expensive fabrication equipment, hindering widespread adoption. This work presents the first end-to-end automated method that generates a fully functional flexible printed circuit board (FPCB)-based tactile glove design directly from a single hand image. Our approach integrates computer vision–driven 3D hand reconstruction, parametric FPCB sensor layout optimization, resistive pressure-sensing modeling, and manufacturability-aware constraints. The pipeline outputs production-ready Gerber files compatible with commercial PCB manufacturers, enabling both personalized fit and scalable manufacturing. Each glove costs ≤$130 and requires <15 minutes of assembly time. Experimental evaluation demonstrates excellent pressure-response linearity and validates user-level reliability and comfort across four prototypes. By eliminating manual design and high-cost tooling, this framework substantially lowers the development barrier for tactile gloves, advancing accessible, low-cost haptic human–machine interfaces.

This work proposes a low-cost, multifunctional sensing fiber for soft robotics, addressing limitations of existing approaches that often rely on expensive materials and complex fabrication processes. The fiber integrates commercially available conductive yarns within silicone tubing and enables dual-modal sensing—resistive strain and capacitive touch/proximity detection—through a simple manual threading process (e.g., 20 cm in under two minutes). Exhibiting high flexibility, ease of fabrication, and reparability, the fiber was successfully integrated into diverse soft robotic systems, including pneumatic grippers for trigger control, flexible bands for pose estimation, soft structures for deformation monitoring, and robotic arms for touch interaction and gesture following. These demonstrations highlight its promising applicability in knitted architectures for wearable and soft robotic applications.

Current robotic systems lack low-cost, scalable, and easily integrable tactile sensing solutions, limiting their performance in fine manipulation and environmental interaction. This work proposes a plug-and-play flexible piezoresistive tactile sensing module featuring a sealed tri-layer composite structure (FPC–Velostat–FPC), enabling high-consistency batch fabrication and excellent mechanical compliance. The module integrates a high-density flexible sensor array, a compact multi-channel readout circuit, low-power serial communication, and flexible printed electrodes, supporting real-time tactile signal transmission at 100 Hz across diverse form factors such as fingertip sensors and large-area tactile pads. Designed for cross-platform compatibility, the system leverages GPU-accelerated tactile simulation and vision–tactile fusion, demonstrating successful applications in contact-aware decision-making, cross-embodiment skill transfer, and sim-to-real fine-tuning tasks.

This study addresses the limitations of existing 3D-printed artificial skins, which are often constrained by single-modality sensing and rigid materials, hindering simultaneous achievement of compliance, high spatial coverage, and multifunctional perception. The work presents the first integration of time-of-flight (ToF) and self-capacitance (SC) dual-modal sensing within a 3D-printed artificial skin, leveraging a flexible composite material and a monolithic electrical interface design. This enables concurrent capabilities in contact detection, proximity sensing, scene reconstruction, and pressure response. The developed skin module comprises 40 sensing units, and six such modules were deployed on an FR3 robotic arm, demonstrating exceptional multimodal perception, conformal adaptability, and impact resistance.

This study addresses the challenge of achieving non-destructive, high-precision, and low-cost thickness control of soft polymeric films during spin coating—a critical limitation for devices such as dielectric elastomer actuators. The authors present a low-cost, desktop-scale spin-coating platform based on 3D printing, integrated for the first time with a low-deformation, laser-reflection-based in situ thickness measurement system. By employing a quadrant photodetector to track real-time displacement of the reflected laser beam, the system enables controlled fabrication of soft films with thicknesses ranging from 50 to 300 µm. Optimized via finite element analysis and calibrated using metal shims, the setup achieves a thickness resolution of 3.6–3.7 µm and a repeatability error of 13 µm (95% confidence interval), successfully producing silicone films with deviations from target thickness of less than 9 µm.

This study addresses the unclear relationship between the structural design of knitted strain sensors and their actual joint motion sensing performance. By integrating uniaxial cyclic tensile testing, simulated joint flexion, and machine learning models, it systematically investigates how knitting structural parameters influence electromechanical metrics and correlate with angle estimation errors. The findings reveal that the gauge factor and baseline resistance are critical predictors of estimation error, whereas other metrics exhibit weak correlations. Furthermore, this work demonstrates that simple tensile tests can effectively screen out suboptimal sensor designs. Ultimately, this research provides a theoretical foundation and an efficient evaluation paradigm for optimizing sensor designs tailored to specific applications.

Existing vision-based methods struggle to achieve reliable, real-time 3D deformation sensing due to line-of-sight occlusions and spatial constraints. This work proposes a skin-inspired flexible sensor that embeds a two-dimensional mirror-stacked array of oxidized eutectic gallium–indium (o-EGaIn) strain gauges within an elastomeric membrane. By integrating off-neutral-axis strain measurement, a mechanics-guided observation model, and a fast optimization algorithm, the system enables high-precision, real-time 3D shape reconstruction without requiring line-of-sight and conforming to arbitrary curved surfaces. A 5×5 sensor array with 12 mm spacing achieves an average reconstruction error of only 0.62 mm and a latency of 0.1 s across diverse deformations. The approach has been successfully demonstrated in applications including gesture recognition, tactile interaction, and intraoperative monitoring, overcoming longstanding limitations of conventional sensing in occluded and spatially confined environments.