This PhD project investigates how data-driven methods can support the transition from traditional time-based inspection practices to risk-based inspection and certification of elevators and escalators. Although these systems generate increasing amounts of operational, maintenance, inspection, and IoT sensor data, inspection regimes are still largely based on fixed schedules and expert judgement.
The research aims to develop methods for equipment health assessment, degradation and failure prediction, reliability estimation, and risk quantification by combining techniques from reliability engineering, machine learning, and predictive analytics. By integrating heterogeneous data sources, the project seeks to improve the understanding of asset condition and future risk evolution. The ultimate objective is to support more effective maintenance, inspection, and certification decisions, enabling safer and more efficient management of elevator and escalator systems. The research also aims to provide a scientific foundation for future risk-based regulatory approaches and is conducted in close collaboration with industrial and certification partners to ensure practical relevance.
