Enabling Tactile Feedback in Robotic Manipulation via FBG Sensors and Haptic Interfaces

Enabling Tactile Feedback in Robotic Manipulation via FBG Sensors and Haptic Interfaces Anna Labardi Biorobotics Inst., Dept.Exc. Rob&AI Sant’Anna School of Advanced Studies Pisa, Italy [email protected] Domenico Prattichizzo DIISM and HHCM University of Siena and IIT Siena, Italy [email protected] Mariangela Filosa Biorobotics Inst., Dept.Exc. Rob&AI Sant’Anna School of Advanced Studies Pisa, Italy [email protected] Gionata Salvietti DIISM and HHCM University of Siena and IIT Siena, Italy [email protected] Tommaso Lisini Baldi DIISM University of Siena Siena, Italy [email protected] Calogero Maria Oddo Biorobotics Inst., Dept.Exc. Rob&AI Sant’Anna School of Advanced Studies Pisa, Italy [email protected] Abstract— The sense of touch is essential for dexterous and safe robotic object manipulation. The same holds in teleoperation, where a major challenge is still how to deliver intuitive tactile feedback from the robot to the human operator. This work presents a wireless, real-time system that integrates an array of Fiber Bragg Grating tactile sensors in a custom robotic gripper, with a wearable haptic interface. Multiple sensor signals were combined and transformed into intuitive force feedback, allowing a remote user to perceive manipulation interactions of the robot with an object in a quasi-natural manner. Keywords—Tactile Sensing, Haptic Interface, Human Robot Interaction I. INTRODUCTION The sense of touch has a fundamental role in human interaction with the environment, enabling fine manipulation, object recognition, and safe physical contact. Translating this capability into robotics has been a longstanding challenge. In recent years, tactile sensors have become increasingly important for enhancing robotic dexterity and autonomy [1]. Beyond the acquisition of tactile information, an equally critical step is the transmission of this information to a human operator in a meaningful way. The integration of tactile sensing with haptic feedback systems enables users to directly perceive contact events or interaction forces experienced by a robot, facilitating more natural and precise control [2]. Haptics has found applications in robotics across a wide range of domains: in telepresence and teleoperation for precision and operator confidence in manipulation tasks in remote environments [2]; in surgical robotics and telepalpation to enable the perception of subtle tissue properties essential for diagnosis and intervention [3]; in industrial and hazardous scenarios for safety via intuitive manipulation in otherwise inaccessible conditions [4]; in rehabilitation and assistive robotics for motor learning and patient engagement, while in virtual and augmented reality for enriching realism and strengthening user embodiment. This work describes a real-time system consisting of the WEART TouchDIVER haptic interface (Fig. 1-a) [5] driven by tactile information acquired through an array of Fiber Bragg Grating (FBG) sensors integrated in a robotic gripper (Fig. 1-b). The proposed system combines signals from multiple FBGs to extract tactile features, which are then transmitted to the haptic interface. While wearing the haptic interface, the user perceives feedback proportional to the interaction forces between the robotic gripper and the environment. This enables the user to experience the sensation of directly pinching the object, creating the illusion that the robotic arm and its end-effector are natural extensions of their own hand and fingers. II. MATERIALS AND METHODS A. Tactile sensing system A robotic manipulator was equipped with a custom gripper consisting of two clamps. Each clamp was covered with a silicone pad embedding an optical fiber integrating 10 FBG sensors (Fig. 1-b). An FBG is a periodic modulation of the effective refractive index in the fiber core. This grating acts as a wavelength-selective mirror, back-reflecting a narrow portion of the transmitted light and generating a reflection peak at the resonance wavelength (𝜆): 𝜆=2 𝑛 𝛤 , () where 𝑛 is the effective refractive index and Γ the grating period, both varying with the applied strain [6]. The fibers were connected to an optical interrogator (FBG-Scan 904, FBGS), to acquire data at 100 Hz and stream them via TCP/IP to a host computer (PC1). A LabVIEW (NI, USA) Graphical User Interface (GUI) was developed to receive, visualize and process the signals before forwarding them to the haptic interface for real-time tactile feedback (Fig. 3). B. Haptic feedback system To provide tactile feedback, the WEART TouchDIVER, a compact and wireless haptic interface designed for extended reality [5], was employed. It features a bracelet-like unit with a hand module and thimbles on the thumb, index, and middle fingers, capable of delivering multimodal feedback (force, temperature, and texture). In this study, only force feedback was considered: each thimble integrates a servomotor that drives a small platform along a guide, exerting pressure on the finger to reproduce the force sensation. The device communicates with a host computer (PC2) via Bluetooth through a dedicated dongle. Its operation relies on a native Fig. 1. Sensing and feedback devices. a) WEART TouchDIVER haptic interface [5]; b) sensorized gripper (highlighted in the red dashed box). 2025 I-RIM Conference October 17-19, Rome, Italy ISBN: 9788894580570 10.5281/zenodo.17629770 139 middleware that manages connection establishment, parameter configuration, and actuator control [7]. Once paired, the PC2 drives the thimble actuators (Fig. 3). C. Processing unit The 20 wavelength data ( 𝜆) from the FBG sensors integrated in the two gripper clamps were streamed to PC1. As a first step, sensor recalibration was performed by subtracting the offset values (𝜆0), so that real-time plots could display the wavelength shifts (Δ𝜆): Δ𝜆𝑘 = 𝜆𝑘−𝜆𝑜,𝑘 ∀ 𝑘∈ℤ,1≤𝑘≤2𝑁 , () where N=10 is the number of sensors per clamp. These recalibrated signals were then processed to extract compact tactile information suitable for haptic rendering. Since one clamp had to be mapped to one thimble of the haptic interface, the sensor values per each clamp were merged by computing their Euclidean norm (∥Δ𝜆∥2). For each clamp (C1, C2), the norm was obtained as: ∥Δ𝜆∥2,𝐶1 𝑜𝑟 𝐶2= √∑Δ𝜆𝐶1 𝑜𝑟 𝐶2,𝑖 2𝑁 i=1 , () This choice integrates the contribution of all the sensors, combining their collective response into a single value per clamp. Then, the resulting norm values were wirelessly transmitted via UDP to PC2, dedicated to controlling the haptic interface. Once received, the values were scaled to the input range required by the WEART SDK, which expects percentages between 0 and 1, corresponding to forces up to 5 N [7]. Accordingly, the normalized inputs for the two thimbles (𝐼𝑇1,𝐼𝑇2) were computed as: 𝐼𝑇1 𝑜𝑟 𝑇2 =∥Δ𝜆∥2,𝐶1 𝑜𝑟 𝐶2 ∥Δ𝜆∥2,𝐶1 𝑜𝑟 𝐶2,𝑀𝐴𝑋 ∙𝐺 , () where ∥Δ𝜆∥2,𝑀𝐴𝑋 is the empirically defined saturation value for each clamp, and G=10 is a gain factor introduced to amplify the signal and perceptual rendering (Fig. 2). III. CONCLUSIONS Integrating tactile sensing with haptic feedback allows users to directly perceive the forces and contact events experienced by a robot, enhancing both precision and operator confidence in object manipulation across a wide range of applications. In this work, a robotic gripper was equipped with multiple FBG sensors, whose signals were processed and transformed into real-time force feedback via the WEART TouchDIVER interface. This fully wireless architecture allows seamless integration between the tactile sensing platform and the haptic interface, ensuring remote, real-time and intuitive delivery of tactile cues to the user. Future studies will integrate neuromorphic touch and haptic feedback [8]. ACKNOWLEDGMENTS This study was supported by the Italian MUR via the ARTIS PRIN (CUP: J53D23000630006). REFERENCES [1] L. Massari et al., “Functional mimicry of Ruffini receptors with fibre Bragg gratings and deep neural networks enables a bioinspired large-area tactile-sensitive skin,” Nat Mach Intell, vol. 4, no. 5, pp. 425–435, 2022. [2] C. Pacchierotti and D. Prattichizzo, “Cutaneous/Tactile Haptic Feedback in Robotic Teleoperation: Motivation, Survey, and Perspectives,” IEEE Transactions on Robotics, vol. 40, pp. 978– 998, 2024. [3] R. V. Patel, S. F. Atashzar, and M. Tavakoli, “Haptic Feedback and Force-Based Teleoperation in Surgical Robotics,” Jul. 01, 2022, Institute of Electrical and Electronics Engineers Inc. [4] C. González, J. E. Solanes, A. Muñoz, L. Gracia, V. Girbés-Juan, and J. Tornero, “Advanced teleoperation and control system for industrial robots based on augmented virtuality and haptic feedback,” J Manuf Syst, vol. 59, pp. 283–298, Apr. 2021. [5] weart, “Weart Touchdiver,” https://www.weart.it/touchdiver/. [6] T. Erdogan, “Fiber Grating Spectra,” JOURNAL OF LIGHTWAVE TECHNOLOGY, vol. 15, no. 8, 1997. [7] Ö. E. Dural, A. A. Shahid, G. Gioioso, D. Prattichizzo, F. Braghin, and L. Roveda, “Evaluation of Tactile Feedback for Teleoperated Glove-Based Interaction Tasks,” in Springer Proceedings in Advanced Robotics, Springer Nature, 2024, pp. 79–93. [8] A. C. Pereira Resende da Costa, M. Filosa, A. Barbosa Soares, and C. M. Oddo, “Type II mechanoreceptors and cuneate spiking neuronal network enable touch localization on a large-area e-skin,” Nat Mach Intell, vol. 7, no. 8, pp. 1278–1291, 2025. Fig. 3. System overview. The gripper sensor data are read through an optical interrogator, processed on PC1 and sent to PC2, where they are appropriately rescaled and given in input to the haptic interface thumb and index thimbles. Fig. 2. Signal processing pipeline. The shown signals are intended only to support visualization and understanding of the processing algorithm, and do not represent actual recorded data. 140