IntelliMan_WP5_Grasping, Manipulation and Arm-Hand Coordination_T5.1_Data fusion and sensing technology_IEEESENSORSLETTERS2024_Data

This dataset contains tactile signals acquired using a biomimetic piezoelectric sensing system during controlled indentation experiments aimed at discriminating different levels of object hardness. The sensing system consists of an array of eight P(VDF-TrFE) piezoelectric sensing units fabricated on a flexible substrate and integrated into a soft, finger-shaped silicone cap. The tactile signals were acquired from eight sensing channels at a sampling frequency of 2 kHz. The dataset was collected using a three-axis Cartesian robot performing indentation experiments on five 3D-printed cubic objects made of thermoplastic polyurethane (TPU), characterized by different infill percentages (3%, 5%, 7%, 10%, and 12%), corresponding to five relative hardness levels. The experimental protocol involved four target indentation loads (8.5, 9.5, 10.5, and 11 N) and three indentation speeds (2.4, 5.6, and 7.2 mm/s). Each combination of object hardness, target load, and indentation speed was repeated 10 times, resulting in a total of 600 trials, with 120 observations per hardness class. Each trial records the piezoelectric response throughout the indentation sequence, with the number of acquired temporal samples varying according to the applied load and indentation speed. The dataset was used to investigate machine learning approaches for object hardness discrimination, including support vector machines (SVMs), single-layer feed-forward neural networks (SLFNNs), and k-nearest neighbors (KNN). Statistical time-domain features were extracted from the tactile signals, and principal component analysis (PCA) was employed to identify the most informative features and reduce the dimensionality of the classification problem. The study also investigated the influence of indentation speed and applied load on classification performance. The dataset is associated with the following peer-reviewed publication, which provides a detailed description of the tactile sensing system, experimental setup, data acquisition methodology, and machine learning classification framework: H. Bassal, Y. Abbass, C. Gianoglio, and M. Valle, "Hardness Discrimination Using Piezoelectric-Based Biomimetic Tactile Sensor and Machine Learning," IEEE Sensors Letters, vol. 8, no. 8, pp. 1–4, Art. no. 5503204, 2024. DOI: https://doi.org/10.1109/LSENS.2024.3431430 Users of this dataset are kindly requested to cite the above publication when using the data in their research.