Robotic Deployment of Embedded Strain Sensors in Precast Tunnel Segments

This article demonstrates the robotic deployment of sensor nodes into precast concrete tunnel segments during manufacturing. Magnetic embeddable sensor nodes based on vibrating wire strain gauges (VWSGs) were deployed on a steel precast segment mold using a six-axis collaborative robot at the lab scale. Robotic sensor deployment proved to be significantly more accurate and consistent than manual sensor deployment methods. On average, positional and angular errors in sensor placement were reduced by 85% when using robotic deployment. The strain transfer coefficients for robotically embedded sensors were evaluated using mechanical bending tests and a finite element model (FEM). Strain transfers across a population of ten segments were found to be $0.93~\pm ~0.012$ in the longitudinal direction, and $0.567~\pm ~0.011$ in the transversal direction. The repeatability of strain measurements within these segments was also confirmed, with low coefficient of variation values of 1% for longitudinal strains and 1.9% for transversal strains. The work presented in this article underscores the measurement performance enhancements that can result from using robotics for sensor deployment in precast manufacturing environments. This could translate to a lower uncertainty and risk for civil asset managers and structural health monitoring (SHM) practitioners.