Dataset of 20 Robot Nudge Gestures Across Affective States and Social Intent

This dataset contains 20 nudge gestures designed by a professional animator who served as the project’s motion expert. Each gesture was authored specifically for a non-humanoid tabletop robot used in research on human–robot interaction. The gesture set is defined by five affective states (excited, content, angry, sad, neutral) crossed with four social intents (taking space, seeking attention, seeking comfort, no intent), resulting in 20 distinct motion specifications.Files follow the naming convention <b>XY.txt</b>, where <b>X</b> denotes the intended social function (S = taking space, A = seeking attention, C = seeking comfort, N = no intent) and <b>Y</b> denotes the affective profile (E = excited, C = content, A = angry, S = sad, N = neutral). Each file contains a time-ordered sequence of robot poses and drive commands intended for execution on a Stewart–Gough platform mounted on an omnidirectional mobile base.Each line in a file consists of a timestamp followed by ten numerical values that describe the robot’s physical pose and base movement at that moment. Platform translations (<b>x_mm, y_mm, z_mm</b>) are expressed in <b>millimetres</b>, and platform rotations (<b>pitch_deg, roll_deg, yaw_deg</b>) in <b>degrees</b>, representing the intended 6-DoF pose of the Stewart–Gough platform relative to a neutral home position. The four additional values (<b>x_drive, y_drive, turnL, turnR</b>) are <b>unitless normalised velocity commands</b> in the range −1.0 to +1.0, specifying the desired lateral motion, forward/back motion, and rotational components for the mobile base. These parameters preserve the animator’s intended timing, rhythm, and expressive qualities and enable precise reproduction of the gestures. full specification of the custom robotic platform used to deliver these nudges, including hardware design and control software, is available through the GitHub link provided in the associated research paper.These expert-designed gesture specifications were used in a two-part human-subject study investigating how people interpret brief robot-initiated nudges, including perceived affect, intent, and agency. The dataset supports research on expressive movement, robot-initiated touch, haptic communication, and non-anthropomorphic social robots.