Current studies conducted in the field of robotic-assisted spinal surgery are focused on using navigation as a tool for optimizing pedicle screw placement and some robotic solutions are already marketed [1], [2], [3]. However, these developments primarily concentrate on outlining and securing the trajectory. The next advancement in robotic-assisted spinal surgery involves optimizing the drilling process itself. Previous studies conducted by the medical device company SpineGuard, in collaboration with the ISIR robotic laboratory, laid the groundwork for implementing a robotic drilling process including automatic breach detection using measurement of tissue conductivity at the tip of the drilling instrument [4], [5]. This paper aims to extend this research by improving the methodology and demonstrating its performance in a more clinically relevant setup. It presents the outcome of 50 drillings performed in pig vertebrae, using an improved breach detection algorithm applied to a trajectory going through the pedicle towards the spinal canal.