Implementation of satellite geodetic GNSS methods and inertial navigation systems for automated geocontrol of construction processes in dense urban environments

Abstract

The article is devoted to the implementation of integrated GNSS/INS navigation systems for automated geocontrol of construction processes in dense urban environments, where the urban canyon effect, multipath interference, and electromagnetic disturbances significantly reduce the reliability of satellite positioning. The aim of the study is to develop and experimentally validate a robust methodology for integrating GNSS and inertial navigation system data to ensure continuous and sufficiently accurate coordinate support for automated construction machinery control systems and digital twins of construction production. The materials and methods include field experiments at high-rise construction sites using multisystem dual-frequency GNSS receivers (GPS, GLONASS, Galileo, BeiDou) integrated with MEMS-based IMUs installed on 12 units of construction machinery (bulldozers, graders, excavators, asphalt pavers); processing of more than 4 TB of raw data was carried out in an extended Kalman filter environment implementing both loosely and tightly coupled integration schemes, followed by verification using robotic total stations, PPK solutions, and terrestrial laser scanning. It is shown that the use of the hybrid GNSS/INS system reduces the RMS of horizontal coordinates under partial horizon obstruction from 4.7–5.1 to 1.5–1.6 cm, and under deep canyon conditions from 18–21 to 3–3.5 cm, while increasing solution availability from ~42 to >94 %, thus ensuring compliance with construction tolerances. The inertial subsystem drift was analyzed, revealing that with signal outages up to 10–20 s, the error does not exceed 2–3 cm, and the filter reinitialization time is less than 1.5 s. The production analysis demonstrates an increase in equipment productivity by 40–52 %, a reduction in rework share to less than 2 %, and fuel savings of up to 18–19 % due to fewer passes and higher first-pass accuracy. The prospects for integrating the proposed algorithms with BIM models, collaborative navigation technologies, and artificial intelligence elements for predictive assessment of the navigation field quality and transition to fully developed digital twins of construction processes are discussed.