Mar Baselios College of Engineering and Technology (Autonomous), is an engineering educational institution located at Thiruvananthapuram, Kerala, India offering engineering education and research. The college is located on a hillock in the Bethany Hills. The educational Institution is situated along the way from Kesavadasapuram to Mannanthala route, this road further extends to north of Kerala as the MC Road.The college is a part of the Mar Ivanios Vidyanagar Campus which has 22 educational institutes, including primary, secondary and higher secondary schools, training institutes and an arts college. The college which started operations in July 2002 is affiliated to the APJ Abdul Kalam Technological University.It is one of the top ranked colleges in Kerala for engineering. All B.Tech. programmes have been accredited by the National Board of Accreditation w.e.f 1 July 2016.Tech.Tech.Tech.Tech.Tech.
This study aims to design, fabricate, and evaluate a portable incubator tailored for onsite microbiological testing in low-resource and remote environments. The primary objective is to address the challenge of maintaining stable incubation temperatures, crucial for accurate microbial analysis in the field. The incubator incorporates an ESP32, a DHT22 temperature sensor, and a 50 W DC PTC heater, integrated with proportional–integral–derivative (PID) control to ensure temperature stability within ± 0.5 °C. The system supports real-time monitoring via Bluetooth, and the power unit converts 220 V AC to the required DC voltages. Performance tests were conducted to assess the incubator’s ability to maintain stable temperatures of 37 ± 0.18 °C, ensuring reliable microbial growth. The incubator demonstrated stable temperature regulation, yielding microbial growth results comparable to standard laboratory incubators. Field tests in Cambodia confirmed its effectiveness in detecting microbial contamination, including Escherichia coli. Experimental validation (n = 48 trials) demonstrated temperature stability of 37.0 ± 0.18 °C (mean ± SD) over 24 h, with a rise time of 4.2 min, a settling time of 8.7 min, and an overshoot < 1.5
All-Optical Orthogonal Frequency Division Multiplexing (AO-OFDM) is a promising technique for high-capacity optical communication. However, its practical application is constrained by a high peak-to-average power ratio (PAPR). This paper describes a 100 Gbps All-Optical OFDM architecture for a radio-over-fibre system that uses optical couplers and phase shifters to realise various optical transforms. A phase pre-emphasis technique is used to reduce PAPR. The performance of DHWT, DFT, DHT, and DCT is evaluated in terms of PAPR and BER versus OSNR. Simulation results show that phase pre-emphasis significantly reduces PAPR by 3–4 dB while maintaining acceptable BER performance, and DHWT-based AO-OFDM outperforms other transforms. At the CCDF of 10^ - 3 the proposed DHWT had a PAPR of around 8.2 dB, which was lower than the existing 9.6 dB DFT-based AO-OFDM. These results confirm that the proposed DHWT-based AO-OFDM has a lower PAPR than the existing DFT, indicating increased power efficiency. Furthermore, for N = 16 subcarriers, both DHWT and DFT require 64 couplers and 64 phase shifters, while DHT and DCT eliminate optical phase shifters, retaining the same number of optical couplers, resulting in lower hardware complexity and insertion loss.
Voice-language pathologists experience severe challenges owing to the lack of a good speech recognition system. This situation makes speech assessment and therapy time-consuming and error-prone. To address these challenges, this research proposes a quantum-enhanced attention-integrated Convolution Neural Network (CNN) model for clinical Malayalam speech recognition that streamlines the speech assessment process. The system is particularly useful for those with hearing or speech impairments because it provides automatic speech recognition, real-time feedback, and structured speech training, thereby assisting pathologists to provide more effective and timely interventions. The proposed model integrates classical and quantum computing principles. Initially, spectrogram features are extracted from the speech samples. These features are fed into the classical convolutional neural network to extract speech features. These extracted features are then passed into the attention mechanism to highlight the most discriminative features. The selected features are encoded into a quantum circuit, enabling quantum-enhanced feature transformation. The resulting quantum embeddings are then fused back with the classical representation using a gated residual mechanism, forming a unified classifier. For dataset creation, 17 words were selected from the standard word list developed by the All-India Institute of Speech and Hearing (AIISH), Mysore. Speech-language pathologists commonly use this word list during speech therapy sessions for assessment and training purposes. In this study, a speech dataset comprising 1,699 utterances was collected. The proposed model achieved an average accuracy of $95.82 \%$ compared to $92.8 \%$ accuracy for the classical model.
Metal additive manufacturing (AM) has garnered significant interest since its emergence owing to its economic efficiency and exceptional design versatility. In contrast to conventional manufacturing methods like welding and casting, additive manufacturing is anticipated to embark on a distinctive evolutionary path, propelled by its specific capabilities and technical framework. The possibilities for innovation and performance improvement through AM much exceeds that of traditional methods. The current research provides an in-depth review of metal AM, a rapidly advancing domain characterized by sophisticated technology and methodologies. The main goal is to provide a comprehensive review of process parameters, challenges and defects, new research advancements, technological developments, and the relationships between process, structure, and properties in metal additive manufacturing. Significant focus is placed on innovative approaches aimed at improving material properties, while simultaneously addressing critical challenges, and application-specific requirements. This paper ultimately addresses recent advancements, existing constraints, and prospective trajectories in metal additive manufacturing.
Early identification of speech and language disorders in children is critical for effective clinical intervention and healthy cognitive development. Conventional assessment methods are largely manual, time-consuming, and subject to evaluator inconsistencies. This paper presents a Digital Speech-Language Assessment App with Adaptive Evaluation and Testing, a bilingual mobile-based platform designed to digitize and standardize pediatric speech and language assessment. The system is developed using Flutter for cross-platform deployment and Firebase for secure authentication, real-time data storage, and synchronization. The application supports bilingual assessment in English and Malayalam, making it suitable for multilingual regions. An age-adaptive testing framework dynamically recommends appropriate reception and expression tests based on the child's age and historical performance. Both automatic and therapistassisted manual scoring are supported to ensure clinical flexibility and accuracy. A Language Proficiency Progress Tracker visually monitors long-term developmental trends to assist in personalized therapy planning. The system is designed with strict data privacy and role-based access control to meet healthcare data standards. Initial implementation and testing validate secure therapist authentication, profile management, and age-based test assignment. The proposed system improves assessment efficiency, diagnostic consistency, and accessibility for pediatric speech and language evaluation.
