MEMS-embedded smart segments for shield tunnel monitoring from manufacture to operation

Traditional shield tunnel monitoring faces critical limitations involving labor-intensive sensor installation, signal instability, delayed deployment, and an inability to track structural behavior throughout the full-lifecycle starting from the manufacturing phase. An intelligent monitoring system comprising smart segments, a wireless gateway, and a cloud platform was developed to enable continuous data acquisition from casting through operational phases. Building on finite element analysis of ten single-ring assembly scenarios and a horizontal convergence calculation method derived from segment rotational displacements, the optimal wireless tilt sensor embedment positions were identified as 271.8° for block L1 and 309.6° for block B1, the former obtained directly from 4.7 million Monte Carlo simulations addressing surface surcharge uncertainty and the latter from the Monte Carlo–characterized rotational-displacement ratio between the two blocks. By embedding micro-electro-mechanical systems (MEMS) wireless tilt sensors into reinforcement cages before casting, this study realized the first full-lifecycle implementation of smart segments. Continuous full-lifecycle monitoring spanning nearly 17 months from segment casting through tunnel operation has validated the system performance, with structural deformation analysis covering 15 months from post-assembly through the operational phase. Temperature data effectively captured cement hydration peaks exceeding 59 ℃, with the temperature later stabilizing at about 24 ℃ in the underground environment. Battery voltage readings demonstrated low power consumption with a net decrease of up to 1.7% after 500 days, yielding an estimated battery life exceeding 15 years. Calculated horizontal convergence agreed with a single total station measurement at the corresponding time and later stabilized at about 23.1 mm. Additionally, data analysis revealed that shield tunneling causes greater fluctuations in the tilt angles of the existing segments than the subsequent segment assembly does. This preliminary application shows that smart segments with embedded sensors can operate reliably at low power and can be deployed without the construction disruption associated with conventional sensor installation.