On-demand vision-based swing suppression for offshore cranes: An onshore field study

Residual payload pendulation on offshore cranes can arise from diverse sources: load pickup, slewing, residual motion-compensation error and unmodeled structural dynamics that must be arrested before final placement. This paper presents an on-demand, vision-based swing suppression system that activates on operator request. A tip-mounted camera coupled with an inertial measurement unit (IMU) detects a hook-mounted marker, and a lightweight pendulum state estimator fuses geometry-based propagation with filtered vision updates at 200 Hz, achieving > 99% detection availability across four field sessions ( ≈ 4 h of active operation), to recover the payload’s swing amplitude, phase, and energy partition. A phase-aware planner schedules a single, quarter-period velocity pulse toward the next swing apex. Onshore field trials on a full-scale motion-compensated crane (Seaonics ECMC C25) with two payloads (500 kg and 1500 kg) across 51 maneuvers yielded a median sway amplitude reduction of 12.8 dB (77%). The method is positioned relative to input shaping and continuous energy damping to clarify its role among existing approaches.