Marthandam College of Engineering & Technology (MACET) is a Private Engineering College located in Kuttakuzhi, Marthandam, Kanyakumari District, Tamil Nadu, India. It was founded in September 2006. Marthandam College of Engineering and Technology (MACET) is the fruit of efforts by the members of Marthandam Educational & Charitable Trust (MECT). The college was approved by All India Council for Technical Education (AICTE) and affiliated with Anna University, Chennai..
Pipeline infrastructure is increasingly vulnerable to complex, multi-stage corrosion defects that compromise structural integrity, operational reliability, regulatory compliance, and long-term cost-effectiveness. Conventional periodic maintenance strategies often lack adaptability to real-time degradation dynamics and environmental variability, resulting in inefficient resource utilization and elevated failure risks. This paper introduces ARPC-DOX (Advanced Risk-Informed ABCDE prioritization and evolutionary cost-optimal framework for intelligent Corrosion Maintenance in Pipelines), a novel AI-driven decision-support framework that integrates predictive modeling, dynamic risk assessment, and cost-aware optimization for intelligent pipeline maintenance. ARPC-DOX fuses a cross-Bayesian network, augmented with cross-attention mechanisms, to capture complex spatiotemporal dependencies in corrosion progression and detect evolving multi-stage defect patterns with high precision. The framework consolidates heterogeneous data sources including inspection logs, sensor telemetry, operational parameters, material characteristics, and environmental stressors into a unified, context-aware predictive engine. Maintenance prioritization is performed through an enhanced ABCDE framework, which is embedded with a dynamic weight adjustment layer that adaptively recalibrates risk weights based on current degradation trends and uncertainty profiles. To optimize the trade-off between maintenance cost and failure risk, the system employs the ADDAX algorithm, a robust, adaptive differential evolution-based metaheuristic designed for dynamic, high-dimensional optimization tasks. Extensive simulation studies conducted on a 100-segment synthetic pipeline network demonstrate the framework’s effectiveness in reducing failure probability, enhancing risk responsiveness, and achieving superior resource allocation. ARPC-DOX represents a scalable, intelligent, and real-time corrosion management paradigm that significantly enhances the safety, resilience, and sustainability of modern pipeline systems.
The longevity of superhydrophobic anti-icing coatings is often plagued by mechanical wear, which causes degradation and reduces ice-phobic properties. In this research, a supercritical fluid (SCF)-assisted fabrication method is used to create a hierarchically structured multiwalled carbon nanotube (MWCNTs) and carbon nanofiber (CNFs) composite coating with enhanced nanofiller dispersion and interface stability. The SCF method enables the creation of a strong and networked micro/nano carbon structure, achieving high superhydrophobicity with a water contact angle of 163.2 degrees and a sliding angle of 5.2 degrees. Quantitative shear tests at -20 degrees C show an extremely low ice adhesion strength of 32 +/- 6 kPa, which is greater than a 90% reduction compared to the pristine FRP substrates. Importantly, after block-on-ring abrasive wear tests, the coating maintains a contact angle of 147.2 degrees and a low ice adhesion strength of about 41 kPa, demonstrating durability-coupled icephobic properties. Surface topography analysis, roughness measurements, and Raman spectroscopy (ID/IG = 0.824) confirm the sustained hierarchical architecture and graphitic structure. These results collectively provide a scalable SCF-enabled materials design platform for developing mechanically durable anti-icing coatings with high potential for advanced structural and energy applications.
Solar-thermal systems rely on high-performance photothermal absorbers to convert sunlight into heat. Conventional absorbers are limited by narrowband absorption, poor high-temperature stability, restricted angular tolerance, and manufacturing constraints. This study presents a graphene-based multi-resonator absorber integrating tungsten (W), zirconium (Zr), and SiO2, designed using material selection and machine learning-assisted optimization. The structure includes graphene-coated square resonators, concentric tungsten square rings, zirconium rectangular auxiliary units, and an outer zirconium square ring on a SiO2 substrate, achieving broadband absorption from the ultraviolet (UV) to far-infrared (FIR) spectral regions. Simulated results show near-unity absorption (>99%) for incident angles of 0 degrees-40 degrees in both transverse magnetic (TM) and transverse electric (TE) polarizations. Maximum absorption reaches 99.997%, and an effective absorption bandwidth of approximately 3000 nm (A >= 90%). Machine learning models trained on angle-dependent data predict the absorber's response with R-2 up to 0.9935, guiding optimization of geometric and material parameters. The observed performance improvements arise from the synergistic interplay of plasmonic, photonic, and phononic interactions, leading to enhanced electromagnetic field confinement and effective impedance matching. The proposed design exhibits polarization insensitivity, strong thermal stability, and compatibility with scalable fabrication processes, making it well suited for applications in solar energy harvesting, thermal management, and infrared sensing, thereby contributing to renewable energy and sustainable energy technologies aligned with SDG 7 (Affordable and Clean Energy).
The adoption of brushless direct current (BLDC) motor is a potential solution and is well suited for sustainable transportation applications, where energy efficiency and renewable power integration are critical. In this research, a photovoltaic (PV)-powered BLDC motor is proposed with the introduction of a novel converter and control strategy. The low voltage generated by PV system is improved utilizing quadratic high-gain boost (QHGB) converter, which steps up the voltage from PV to meet the requirement of BLDC motor. The converter’s operation is managed by an improved flower pollination algorithm (IFPA)-optimized proportional–integral (PI) controller, which dynamically adjusts the duty cycle based on real-time voltage requirements. This innovative approach not only enhances power conversion efficiency but also ensures a stable DC link voltage, critical for efficient motor drive operation. Furthermore, to ensure consistent supply of power availability, bidirectional DC–DC converter is incorporated into the system. This bidirectional converter facilitates both charging and discharging of the battery, enabling continuous motor operation and enhancing the system’s reliability. The validation proposed system is examined using MATLAB, and the results demonstrate that the proposed converter ranks with higher efficiency of 97.56
This is an overview of research papers in between the year 2002 to 2024 analyses a mock-up for an AI integrated language learning system to aid students academic and proficiency focused test preparation. The system is designed to be a web-based platform that includes NLP and speech processing, as well as adaptive learning techniques. The LLM environment of such portals is intended to offer grammar and spelling checks, interactive discussion-place training, listening-based exercises, and reading-comprehension units according to standardized (language-test) models. The architecture will integrate a multi model NLP pipeline, speech-to-text processing, and AI-driven feedback to facilitate personalized learning. Since it is a preliminary concept tool, the main interest is in its design and working methodology and the consequences of an ultimately implementation. The platform is expected to improve language proficiency through real-time guidance, automated evaluation, and progress visualization.
