As robots become more integrated into humans' daily activities, it is essential to understand how human decision varies during co-learning with robots in real-world scenarios. Despite great advances in developing humanoid robots, which aims to foster a seamless collaborative world where humans and robots coexist, a gap remains in the social bond between humans and robots, particularly in tasks demanding optimal teamwork. In alignment with current pioneering efforts in the human-robot collaboration field, this paper presents an experimental study leading to a rationale analysis and classification of human behavioral dynamics during a joint collaborative pick-and-place task with a robotic arm. Our post-experimental analysis categorized human behavioral dynamics into three distinct broad categories, which are "strategic explorers and decoders", "reactive navigators and dynamic responders", and "score maximizers and ideal collaborators". We provide in-depth analysis for each group, exploring potential reasons for their observed behavioral patterns and irrational decisions substantiated by intuitions from psychological and behavioral game theory, including concepts of false belief and strategy development.
This study develops an explicit Listing-consistent reduced-order dynamics and nonlinear feedback control framework for generating biomimetic eye movements. Building on the established relationship between Listing’s Law and the Half-Angle Rule, we give a coordinate-specific necessary-and-sufficient proof for trajectories starting from a compatible orientation within a nonsingular gaze domain. A single holonomic constraint and its velocity relation reduce a spherical-eye model from three rotational degrees of freedom and six states to two rotational degrees of freedom and four states. A fixed-frame torque compatibility condition and an explicit inverse control map convert desired yaw and pitch accelerations into two independent torque components and reconstruct the required torsional component. Human saccade recordings inform representative controller gains, which are held fixed across six trajectory comparisons using measured target displacements. These limited comparisons provide a preliminary qualitative assessment of the framework’s ability to generate saccade-like responses. Torsion is determined by Listing’s Law rather than measured independently.