A review on condition monitoring and control of industrial drilling systems: fault detection, fuzzy control, and human-induced faults

This review presents a structured synthesis of recent advances in condition monitoring, fault diagnosis, and control of industrial drilling systems, with emphasis on rotary and rotary-percussive drilling. The paper is guided by a transparent review methodology and positions itself against recent review literature to clarify its scope and contribution. The reviewed studies are organized around four main themes: (i) fault classification across machine, rock-induced, human-induced, and environmental sources; (ii) condition-monitoring and fault-diagnosis methods based on multi-sensor measurements, signal processing, observer-based estimation, and data-driven intelligence; (iii) mathematical modelling of drill-string dynamics, stress-wave propagation, drill–rock interaction, and energy-transfer behaviour; and (iv) comparative evaluation of conventional, observer-based, machine-learning, fuzzy, and clustering-based control frameworks. Particular attention is given to recent developments in multivariate monitoring, wavelet-based anomaly localization, neural-network-assisted diagnosis, digital twins, and adaptive fuzzy control. A key conclusion of the review is that drilling anomalies cannot be interpreted reliably through a machine-only perspective; instead, they should be understood through a coupled machine-rock-human-environment framework. The paper further argues that future intelligent drilling systems should integrate signal fusion, mathematical modelling, hidden-state estimation, and human-aware supervisory logic, with hybrid architectures combining physics-based models, multivariate monitoring, machine learning, fuzzy reasoning, and digital twins representing the most promising pathway toward robust, interpretable, and industrially deployable drilling intelligence.