International Journal of Innovations in Science & Technology
Not a member yet
813 research outputs found
Sort by
Cysteine-Coated Cadmium Sulfide Nanoparticles Conjugated with Curcumin for Antimicrobial Activity
Full text linkNanoparticles have several applications in drug delivery. Attaching therapeutics to specially designed carriers enables precise delivery to specific cells. Nanostructures have unique physicochemical and biological features, such as an increased reactive surface area and the ability to pass through tissues and cell walls due to their small size, making them a promising material for biomedical applications. Cysteine-coated cadmium sulfide nanoparticles were prepared using a wet process at high pressure and temperature, followed by curcumin conjugation. The antioxidant, anticarcinogenic, and anti-inflammatory properties of curcumin are well acknowledged. Cadmium sulfide nanoparticles are of extremely good semiconducting material that shows fluorescence at a particular wavelength in spectrophotometric analysis. X-ray diffraction (XRD) of the nanocomposite was conducted to verify the crystalline nature of nanoparticles and to find the average crystallite size of cadmium sulfide nanoparticles. The Fourier-transform infrared spectroscopy (FTIR) confirmed the conjugation of cysteine with CdS and curcumin. Antibacterial activity of the synthesized material against Escherichia coli (E. coli) cells was assessed at different concentrations. The antibacterial activity of conjugated cadmium sulfide nanoparticles against E. coli bacteria was examined using the well diffusion method. The results showed that cadmium sulfide nanoparticles coated with cysteine and conjugated with curcumin had better cytotoxicity against bacterial infections caused by E. coli bacteria
IoT-Enabled Assistive Glove for Real-Time Sign Language Translation Using Machine Learning
Full text linkThis paper presents a real-time system for translating gestures from American Sign Language (ASL) using an IoT-enabled smart glove. The glove is equipped with five flex sensors and an MPU-6050 gyroscope to capture finger movements and wrist orientation, processed by an Arduino Nano. Sensor data is transmitted via a Bluetooth module to a mobile application, where a Random Forest machine learning model with 97% accuracy classifies the gestures. The recognized gestures are displayed as text and vocalized through a speaker. Moreover, the app has a feature that allows users to see ASL signs with its corresponding vocabulary, thus enabling accessibility and making language more accessible to learn. It enhances the communication between the deaf and the hearing community since it offers an accurate, portable, and interactive sign recognition application
Evaluation And Mitigation of Industrial Fire Hazards in The Faisalabad Industrial Estate Development and Management Company (Fiedmc) Zones
Full text linkIndustrial fire hazards are a major threat to lives, infrastructure, and economic activities, especially in growing urban areas. Although industries significantly contribute to the national industrial sector but it may easily catch fire, resulting in a brutal impact on infrastructure, workers, and the environment. This research seeks to assess the characteristics, causes, and frequency of industrial fires within FIEDMC zones, pinpoint the most frequent ignition sources, and propose effective mitigation measures. A quantitative methodology was utilized to gather fire incident data from 2014 to April 2025, drawing from official reports from rescue services and fire stations on-site. The data was analyzed using Power BI to uncover trends in incidents, injuries, and fatalities, while also identifying the most prevalent causes of industrial fires. Factors considered included overloaded wiring, HVAC malfunctions, human error, structural failures, and boiler issues. The visualizations enabled the categorization of causes and the identification of high-risk years and emerging patterns. The results indicated a notable increase in fire incidents and casualties in recent years, particularly from 2020 to 2025. Electrical and mechanical failures were identified as the primary causes, with overloaded wiring alone contributing to 30.43% of fire cases, followed by HVAC problems (18.84%) and human negligence (15.94%). The highest numbers of injuries and fatalities occurred in 2023, with 410 injuries and 80 deaths, reflecting a significant lapse in safety. The study concludes that FIEDMC zones are facing a persistent fire safety crisis influenced by ingrained weaknesses in risk management. Tackling this issue requires long-term, comprehensive solutions, which include regular inspections, worker training, infrastructure improvements, and stricter enforcement of regulations. To reduce risks related to industrial fire hazard require quarterly safety audits, load assessments of all wiring systems, mandatory fire-response training for staff, and the creation of a dedicated fire risk monitoring unit in each estate. Importantly, fire safety must be treated as a management issue, not just a technical one
Green Growth: AI-Driven Intelligent Farming for Effective Resource Management
Full text linkEffective fertilizer management plays a critical role in maximizing crop yield while reducing environmental harm and minimizing resource waste. This study presents an IoT-based intelligent fertilizer recommendation system designed to deliver accurate, real-time application guidance. The system integrates NPK sensors for soil nutrient detection, environmental sensors for humidity and temperature monitoring, and rain gauges to collect precipitation data. Data from the field is transmitted through an Arduino microcontroller to a cloud platform. A Random Forest classifier is used to determine the need for fertilization, while a CatBoost regressor estimates the required fertilizer quantity. The system was tested using real-time field data across 22 crop types, achieving 100% accuracy in classification and strong performance in regression tasks. Recommendations are automated and delivered via SMS to streamline field operations. The objective of this study is to develop an automated, sensor-driven fertilizer recommendation system using machine learning for precision agriculture. The novelty lies in the integration of real-time IoT sensing with hybrid AI models to optimize fertilizer use. This approach enhances productivity, reduces input waste, and supports environmentally sustainable farming
Adaptive Student Assessment Method for Teaching Programming Course
Full text linkComputer programming is a core component of computer science education and is widely recognized as a vital skill for aspiring professionals. Repetitive coding assessments help students improve their programming abilities, but the manual creation and evaluation of these assessments can be time-consuming and challenging for instructors. To address this, we developed an Adaptive Student Assessment System (ASAS) that automatically generates subjective programming questions aligned with Course Learning Objectives (CLOs) and assists in evaluating student responses. The system was evaluated using a controlled study involving two groups: a test group and a control group. Results demonstrated that the test group consistently outperformed the control group across cognitive assessments, with overall performance improvements of 13.5%. Affective feedback collected through a post-term survey showed a 48.20% higher agreement rate in the test group regarding motivation, clarity, and satisfaction with the assessment process. Teacher evaluations further confirmed the system\u27s effectiveness, with improvements of 23.33% in assessment creation, 26.67% in assessment conduction, and 43.33% in result compilation compared to traditional methods. Teachers reported reduced workload, increased efficiency, and a positive attitude toward long-term adoption of the system. These findings highlight that ASAS not only enhances student engagement and academic performance but also improves instructional efficiency, making it a scalable and effective solution for programming education
Deep Learning Based Medicinal Plant Identification for Enhanced Botanical Conservation
Full text linkPlants used in medicine are an essential part of the human health system with various natural medicines and health properties. The right identification of medicinal plants will support the conservation of these natural resources and enhancement of these traditional medical practices. Medicinal plants can now be identified and classified more precisely and reliably by use of leaf and plant pictures using the technology of artificial intelligence and machine learning, especially deep learning. We used Convolutional Neural Networks (CNNs) deep learning models with transfer learning VGG11, ResNet34, and DenseNet121. The novelty of our study is that we combine DenseNet121 with the Multi-Trend Binary Code (MTBC) feature descriptor to perform better and extend features representation. These models have been tested on two benchmark datasets, which include the Indonesian Medicinal Plants Dataset as well as the Indonesian Herb Leaf Dataset. Although all CNN models performed well in terms of accuracy, the proposed hybrid model, DenseNet121+MTBC, performed better than the remaining, attaining its best accuracy of 94.51%, and offering better precision, recall, and F1-score metrics. The results note the usefulness of the combination of the traditional texture descriptors and deep learning features, thus, the synergistic trait of the hybrid approach. The hand-crafted features combined with DenseNet121 give a more effective solution to the repetitive phenomenon of medicinal plant identification than just any CNN. The method offers a convenient and efficient method of alternative relying on conventional methods of identification, offering proficient, exact, and advantageous rapid medication identification of plants
Testing Chatbot Systems using Agentic AI Approach
Full text linkAs large language models (LLMs) become increasingly integrated into real-world applications, robust and scalable evaluation methods are essential to ensure their reliability, safety, and effectiveness. This work introduces an innovative evaluation framework grounded in an agentic AI simulation approach, designed to overcome the limitations of traditional testing methodologies in newly developed chatbots. Unlike conventional methods that depend on static benchmarks or human evaluators, our approach employs autonomous AI agents capable of simulating a wide spectrum of user interactions. Within a controlled multi-agent environment, these evaluator agents interact with the target chatbot using natural language queries specifically designed to probe various functional capabilities, identify edge cases, and uncover potential failure modes. The agentic evaluation methodology systematically assesses the performance of chatbots in multiple dimensions, including task completion efficiency, contextual understanding in dynamic conversations, and adherence to safety and ethical guidelines. By incorporating recent advances in agentic metrics and automated scenario generation, our system produces detailed data-driven performance reports that capture both strengths and vulnerabilities in chatbot behavior. Preliminary results show that this approach not only reveals significantly more edge cases than conventional methods, but also reduces overall evaluation time by approximately 60-70 percent. This work contributes to a scalable, standardized testing paradigm that better aligns theoretical performance indicators with the practical challenges of deploying LLMs in real-world environments
A Federated Framework for Air Quality Prediction in Smart Cities
Full text linkOver the last couple of decades, due to the constant increase in urbanization and industrialization, the concern in terms of air pollution has become a serious issue. In most cities, the pollution in the air is mostly comprised of Nitrogen Dioxide (NO2), Ozone (O3), Carbon Monoxide, and Particulate Matter, all of which can cause serious health issues. There is an emergent need for a system to detect air pollution. This research presents a framework that uses Federated Learning to lessen the communication overhead during the prediction process and ensure data privacy. The research also uses different Machine Learning algorithms, such as Random Forest, Support Vector Machine (SVM), and Logistic Regression, to train and evaluate the research
Voice Cloning and Synthesis Using Deep Learning: A Comprehensive Study
Full text linkThis paper reviews current voice cloning and speech synthesis methods. It focuses on the way that deep learning enhances AI-generated voice synthesis in terms of quality, flexibility, and efficiency. We analyze the top AI models in terms of their significance to virtual assistants, dubbing, and accessibility tools: XTTS_v2, Whisper, and Llama 8B. Voice cloning and TTS efforts in Tortoise are improved by XTTs_v2. Based on the multilingual creative transfer, it has a higher speed and shorter time of a computational process, and generates synthetic speech closer to naturalness. Whisper is a transcription model that goes from an audio waveform to text. It simplifies access to audio data. Llama 8B focuses on user question answering for enhancing AI and human interaction. Other related work includes fastSpeech2 [1], Neural Voice Cloning with few Samples [2], and Deep Learning-Based Expressive Speech Synthesis [3], which also contribute to these advancements. This progress enhances machines\u27 ability to communicate in an emotional and human-like way, leading to more sophisticated technology
An Efficient Read and Mark Mechanism for Multiple-choice Questions Using Optical Character Recognition
Full text linkThis research paper focuses on modifying the grading of multiple-choice questions (MCQs) to better the efficiency and incorrectness of educational tests. Conventional grading systems, such as optical mark recognition (OMR), have fundamental drawbacks, excluding the necessity for precise shading, time-wasting, and the use of special OMR sheets and OMR scanners. This conceptualization can be expensive and error-prone, especially if the MCQs papers are folded or unmarked. In comparison, the suggested OCR-based approach gives fundamental benefits in all necessary areas. It is less costly to use a simple scanner and software alternatively to costly OMR equipment. The method is motivated to be simple to set up and use. It importantly reduces error rates and marking time by employing precise OCR algorithms and processing greater amounts of answer sheets quickly. Moreover, the system is extremely accurate and scalable, allowing it to handle a rising amount of paper efficiently. It also has limited trust in external tools and is highly flexible and adaptable to different MCQ formats and grading settings. In General, the OCR-based approach outperforms existing methods by eliminating their shortcomings and delivering a trustworthy, time-saving alternative for automated MCQ grading