Predictive maintenance, as a form of pro-active maintenance, has increasing usage and shows significant superiority over the corrective and preventive maintenance. However, conventional methods of predictive maintenance have noteworthy limitations in maintenance optimization and reliability improvement. In the last two decades, machine learning has flourished and overcome many inherent flaws of conventional maintenance prediction methods. Meanwhile, machine learning displays unprecedented predictive power in maintenance prediction and optimization. This paper compares the features of corrective, preventive, and predictive maintenance, examines the conventional approaches to predictive maintenance, and analyzes their drawbacks. Subsequently, this paper explores the driving forces, and advantages of machine learning over conventional solutions in predictive maintenance. Specifically, this paper reviews popular supervised learning and reinforcement learning algorithms and the associated typical applications in predictive maintenance. Furthermore, this paper summarizes the four critical steps of machine learning applications in maintenance prediction. Finally, the author proposes the future researches concerning how to utilize machine learning to optimize maintenance prediction and planning, improve equipment reliability, and achieve the best possible benefit.

Stress wave-based communication has great potential for succeeding in subsea environments where many conventional methods would otherwise face excessive difficulty, and it can benefit logging well by using the drill string as a conduit for stress wave propagation. To achieve stress wave communication, a new stress wave-based pulse position modulation (PPM) communication system is designed and implemented to transmit data through pipeline structures with the help of piezoceramic transducers. This system consists of both hardware and software components. The hardware is composed of a piezoceramic transducer that can generate powerful stress waves travelling along a pipeline, upon touching, and a PPM signal generator that drives the piezoceramic transducer. Once the transducer is in contact with a pipeline surface, the generator integrated with an amplifier is utilized to excite the piezoceramic transducer with a voltage signal that is modulated to encode the information. The resulting vibrations of the transducer generates stress waves that propagate throughout the pipeline. Meanwhile, piezoceramic sensors mounted on the pipeline convert the stress waves to electric signals and the signal can be demodulated. In order to enable the encoding and decoding of information in the stress wave, a PPM-based communication protocol was integrated into the software system. A verification experiment demonstrates the functionality of the developed system for stress wave communication using piezoceramic transducers and the result shows that the data transmission speed of this new communication system can reach 67 bits per second (bps).

In the challenging downhole environment, drilling tools are normally subject to high temperature, severe vibration, and other harsh operation conditions. The drilling activities generate massive field data, namely field reliability big data (FRBD), which includes downhole operation, environment, failure, degradation, and dynamic data. Field reliability big data has large size, high variety, and extreme complexity. FRBD presents abundant opportunities and great challenges for drilling tool reliability analytics. Consequently, as one of the key factors to affect drilling tool reliability, the downhole vibration factor plays an essential role in the reliability analytics based on FRBD. This paper reviews the important parameters of downhole drilling operations, examines the mode, physical and reliability impact of downhole vibration, and presents the features of reliability big data analytics. Specifically, this paper explores the application of vibration factor in reliability big data analytics covering tool lifetime/failure prediction, prognostics/diagnostics, condition monitoring (CM), and maintenance planning and optimization. Furthermore, the authors highlight the future research about how to better apply the downhole vibration factor in reliability big data analytics to further improve tool reliability and optimize maintenance planning.

The proliferation of Internet of Things (IoT) devices has introduced significant security challenges. Resource-constrained devices face sophisticated threats but lack the computational capacity for advanced security analysis. This study investigates optimal security task allocation in Cognitive IoT (CIoT) networks. It specifically examines when IoT devices should process security tasks locally or offload them to Mobile Edge Computing (MEC) servers. The problem is formulated as a Continuous-Time Markov Decision Process (CTMDP). The study demonstrates that the optimal offloading policy has a threshold structure. Security tasks are offloaded to MEC servers when the offloading queue length is below a critical threshold, $k^{\ast }$. Otherwise, tasks are processed locally. This structural property is robust to changes in MEC server configurations and threat arrival patterns. It ensures an optimal and easily implementable security policy under the exponential model. Theoretical analysis establishes upper bounds on the performance of AI-based security controllers using the same models. The results also show that standard model-free Q-learning algorithms can recover optimal thresholds without any prior knowledge of the system parameters. Simulations across multiple reinforcement learning architectures, including Q-learning, State–Action–Reward–State–Action (SARSA), and Deep Q-networks (DQN), confirm that all methods converge to the predicted threshold. This empirically validates the analytical findings. The threshold structure remains effective under practical imperfections such as imperfect sensing and parameter estimation errors. Systems maintain 85% to 93% of their optimal performance. This work extends threshold Markov Decision Process (MDP) analysis from classical queuing theory to the context of CIoT security offloading. It provides optimal and practical policies and model-free algorithms for use by resource-constrained devices.

Admission control governs Quality of Service (QoS) in cognitive IoT networks (CIoTNs) in which secondary nodes opportunistically share spectrum with preemptive primary users. The optimal policy is of threshold type, but the threshold table is indexed by the 2N fully occupied channel configurations, so both model-free learning and on-chain enforcement scale poorly in N. We show that this table has low intrinsic complexity: across 55 randomised continuous-time Markov decision process (CTMDP) instances, a monotone step function of the aggregate secondary-user drain rate σ(n) carrying only 1.6–2.3 distinct levels reproduces 81–87% of the 2N table entries exactly, the remainder erring by a single queue slot. We give a quasi-static argument for why σ(n) is the right scalar summary. Using this regularity as an inductive bias, an index-pooled Q-learning read-out reduces discounted policy-value loss by up to 4.4× (1.55% to 0.36% at N=4, six seeds) at the largest sample budget, but adds variance and was worse than the per-configuration baseline in one of eight cells; we therefore also specify a deployment gate and report it as untested. We further report a negative result: projecting learned value differences onto the concave cone guaranteed by the structural theorem is inert at N=4 and yields at most a 0.22 percentage-point gain at N=3, because recovery is limited by configuration coverage rather than by shape violation. Finally, we implement Service Level Agreement (SLA) enforcement as a Solidity contract and measure it on a local Ethereum Virtual Machine. Index compression cuts policy-installation gas by 241.8× at N=16 (48.79M to 0.20M gas) and keeps installation in one transaction, but raises the per-decision cost by 14.4–18.4k gas; it is therefore a feasibility mechanism for large N and for frequently re-committed policies, not a uniform improvement.