Which technique is used to reduce sensitivity to singularities in robot control?

• A. Proportional Control
• B. Damped Least Squares
• C. Model Predictive Control
• D. Purely Kinematic Control

Answer: (B)

When a robot arm nears a singular configuration, the Jacobian becomes nearly ill-conditioned, so small commanded end-effector motions can cause very large joint motions. The technique used to reduce this sensitivity is damped least squares. It modifies the way we solve for joint velocities from a desired end-effector velocity by adding a damping term to the inverse calculation: instead of using the plain pseudoinverse, we compute J^T (J J^T + lambda^2 I)^{-1}. The added lambda^2 I acts like a stabilizer, preventing huge joint movements when J J^T is close to singular. This yields more robust, smooth behavior near singularities, at the cost of some tracking precision that depends on how large the damping factor lambda is. Other methods like simple proportional control don’t address the kinematic conditioning, model predictive control optimizes over a horizon but isn’t specifically about singularity handling, and purely kinematic control doesn’t include a mechanism to tame ill-conditioning.