Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper surveys and benchmarks singularity avoidance techniques in inverse kinematics (IK) for robotic manipulators, covering both classical and learning-based methods. It introduces a taxonomy based on geometric structure and robustness guarantees, and proposes a benchmarking protocol for evaluating IK solvers. Experiments on the Franka Panda robot reveal that pure learning methods struggle with singularity avoidance, while hybrid approaches combining learning and classical techniques achieve significantly better performance.
Pure learning-based inverse kinematics solvers are surprisingly brittle near singularities, failing even on well-conditioned targets and highlighting the need for hybrid approaches that incorporate classical geometric constraints.
Singular configurations cause loss of task-space mobility, unbounded joint velocities, and solver divergence in inverse kinematics (IK) for serial manipulators. No existing survey bridges classical singularity-robust IK with rapidly growing learning-based approaches. We provide a unified treatment spanning Jacobian regularization, Riemannian manipulability tracking, constrained optimization, and modern data-driven paradigms. A systematic taxonomy classifies methods by retained geometric structure and robustness guarantees (formal vs. empirical). We address a critical evaluation gap by proposing a benchmarking protocol and presenting experimental results: 12 IK solvers are evaluated on the Franka Panda under position-only IK across four complementary panels measuring error degradation by condition number, velocity amplification, out-of-distribution robustness, and computational cost. Results show that pure learning methods fail even on well-conditioned targets (MLP: 0% success, approx. 10 mm mean error), while hybrid warm-start architectures - IKFlow (59% to 100%), CycleIK(0% to 98.6%), GGIK (0% to 100%) - rescue learned solvers via classical refinement, with DLS converging from initial errors up to 207 mm. Deeper singularity-regime evaluation is identified as immediate future work.