IntelliMan_WP5_Grasping, Manipulation and Arm-Hand Coordination_T5.1_Data fusion and sensing technology_IEEENEWCAS2025_Data

This dataset contains tactile signals acquired using a P(VDF-TrFE) piezoelectric sensing system during controlled normal indentation experiments. The sensing system consists of an array of eight screen-printed piezoelectric sensing units fabricated on a flexible substrate. The electrical response of a single sensing unit was acquired at a sampling frequency of 2 kHz, while a strain-gauge-based load cell provided the reference contact force measurements. The dataset was collected using a three-axis Cartesian robot equipped with a soft spherical indenter. The experiments involved seven target peak forces (4.5, 5.5, 6.5, 7.5, 8.5, 9.5, and 10.5 N) and seven indentation speeds (2.4, 3.2, 4.02, 4.8, 5.6, 6.4, and 7.2 mm/s). The dataset comprises 490 indentation trials, organized into seven force classes. Each trial records the piezoelectric response throughout the indentation process, from the application of the contact force to the release of the indenter. The number of acquired samples varies across trials depending on the indentation speed and target force. The dataset was used to investigate machine learning approaches for estimating peak normal contact forces from piezoelectric tactile signals. Statistical time-domain features were extracted and analyzed using principal component analysis (PCA), while an artificial neural network (ANN) was developed to estimate the applied peak force. The dataset is associated with the following peer-reviewed conference publication, which provides a detailed description of the sensing system, experimental setup, data acquisition methodology, and machine learning framework: M. Yaacoub, R. Khalifeh, C. Gianoglio, A. Ibrahim, and L. Seminara, "Machine Learning Based Contact Force Estimation Using a PVDF Tactile Sensing System," 2025 23rd IEEE Interregional NEWCAS Conference (NEWCAS), pp. 430–434, 2025. DOI: https://doi.org/10.1109/NEWCAS64648.2025.11107072 Users of this dataset are kindly requested to cite the above publication when using the data in their research.