The dataset incorporates sensor noise and object variability to ensure robustness in testing and validation. The data generation process involves creating 50 unique objects using MATLAB, with each object assigned a randomly generated Center of Gravity (CoG). These objects exhibit diverse shapes, ranging from regular to irregular geometries, and weights spanning from 2 N to 50 N. The CoG positions are distributed within a range of up to 100 mm from the base of the object along the X, Y, and Z axes. This variability in shapes, weights, and CoG positions ensures a comprehensive dataset that simulates real-world scenarios and challenges, providing a solid foundation for training and evaluating the proposed ML models.