A Cutter Vibration Signal Processing Based State Monitoring Method for Improving Operation Safety of Spinal Automatic Drilling System

This manuscript introduces a state monitoring strategy for spinal pedicle drilling to stop the drill bit automatically at an appropriate position. First, the task and challenge of robotic-assisted bone drilling are discussed, and a robotic system is designed to carry out bone drilling experiments. Then, the drill’s handle vibration is acquired via an accelerometer and processed by the short-time fast Fourier transform to obtain harmonic amplitudes. Subsequently, a signal processing program and a state detection algorithm based on harmonic amplitudes are proposed to extract state feature curves and detect drilling states. Finally, the performance of the developed strategy is verified via drilling experiments on animal bones and the remaining thickness measurement of the drilling area. Experimental results demonstrate that the success rate of stopping the drill bit at the appropriate position is 96.29% in automatic drilling mode and 85.19% in manual drilling mode. Moreover, the success rate for avoiding penetration of the inner cortical bone is 100% in these two drilling modes. The proposed strategy could prevent damage to the blood vessel during the spinal pedicle drilling operation.