• Real-time obstacle avoidance and collision-free operation emerged as a core paradigm, integrating artificial potential fields with distributed control and learning-based strategies to cope with moving obstacles. [1], [9], [20]

• Inverse kinematics for redundant manipulators spurred generalized inverse methods, fast closed-loop algorithms, and continuation-method approaches enabling multiple configurations. [2], [5], [6], [19]

• Robust/nonlinear/adaptive control frameworks provided stability and tracking in the presence of disturbances, constraints, and model uncertainties. [3], [7], [15], [17], [12]

• Trajectory planning and dynamic modeling converged on cubic polynomial trajectories and symbolic dynamic modeling (ARM), balancing offline planning with online tracking. [8], [18], [13], [14]

• Calibration and parameter identification formed the backbone for accurate control through forward calibration methods and refined dynamic models. [4], [18], [16]