Sri Lankan Journal of Technology Feed

Combating Moisture Stress in Wheat Through Pre-Sowing Seed Priming

Tue, 11 Aug 2026 00:00:00 GMT

Soil moisture stress causes less germination and poor stand establishment resulting in growth and yield reduction in wheat which can be mitigated by seed priming to some extent. To judge the efficacy of different seed priming agents in enhancing wheat growth and yield under moisture stress an experiment including four irrigation regimes (no irrigation, single irrigation at Crown Root Initiation (CRI) stage, single irrigation at flowering stage and two irrigations at both CRI and flowering stages and six seed priming agents (namely unprimed control, water, CaCl2, KCl, KNO3 and mannitol) was conducted following a split-plot design with three replications. Seed priming exhibited a positive role in enhancing wheat growth and productivity under both irrigated and moisture stress condition. No irrigation resulted in the lowest grain yield while two irrigations at both the critical stages produced the highest grain yield. The maximum grain yield (3.0 t ha-1) was produced by two irrigations coupled with CaCl2 or KCl priming, while no irrigation coupled with no priming produced the minimum grain yield (1.2 t ha-1). Therefore, seed priming preferably with 20000 ppm KCl or CaCl2 may be practiced to alleviate the moisture stress in wheat to some extent.

Hybrid Test Automation Framework for Graphical User Interface

Tue, 11 Aug 2026 00:00:00 GMT

Modern software development is greatly aided by automated software testing that provides continuous feedback on product quality. Software testing is crucial since the system's complexity increases daily with requirement changes and technological advancements. The Graphical User Interface (GUI) is the most typical interface to interact with the software. Around 60% of the overall code is related to GUI in software. Many solutions have been developed in response to the increased demand for User Interface (UI) test automation. Researchers and developers have been working tirelessly to improve upon existing methods. The GUI testing process has been automated in various ways, such as recording and playback tools, keyword-driven methodology, model-based approaches and hybrid models. A small change in the locations of the GUI elements needs many updates in the testing scripts. This paper introduced a Hybrid Test Automation Framework using selenium to overcome challenges in the GUI functions and property changes. The framework was tested in two versions of the e-commerce web application: E-commerce web application with a unique UI object ID, and E-commerce web application with traditional implementation. An e-commerce web application with a unique UI object ID performs better than a traditional implementation.

Long Short-Term Memory-Based Multivariate Forecasting Model for Dengue: A Case Study in Sri Lanka

Tue, 11 Aug 2026 00:00:00 GMT

Dengue fever has been a significant health concern in Sri Lanka since the 1960s. The number of cases has drastically increased over the last decade. Different statistical and classical machine learning models have been proposed to forecast dengue in order to mitigate the disease from reaching its high transmission rate. More recently, the advent of neural networks has improved prognosis in an efficient manner through time forecasting of dengue data using multiple predictor variables. Understanding the strength of long short-term memory (LSTM) models in the past, this specific research has evaluated three variants of LSTM models: unidirectional LSTM, bidirectional LSTM (BiLSTM), and encoder-decoder LSTM, with the aim of predicting dengue occurrences in Sri Lanka. Weather data, including rainfall and mean temperature, were used as predictors while the efficiency of the models was assessed using RMSE. While all three models exhibited relatively good performance, the BiLSTM model significantly outperformed the other two models. This study affirms the use of LSTM for predicting dengue-like vector-borne diseases that are characterized by complex relationships with the predictors.

Artificial Intelligence as a Catalyst for Transforming Education

Tue, 11 Aug 2026 00:00:00 GMT

Artificial Intelligence (AI) holds transformative potential for education, enabling personalized, scalable learning while also risking the amplification of existing inequities. Achieving SDG 4 in the context of digital transformation requires more than the deployment of tools; it demands deliberate strategies to ensure equity, ethics, and relevance. The UN 2030 Agenda positions Quality Education not only as a standalone goal but as a foundation for all SDGs. The study finds that AI can improve learning outcomes and streamline administrative tasks; however, these benefits are contingent upon strategies that address inequities, data privacy, transparency, and human oversight. By analyzing global AI initiatives, policy frameworks, and educational practices, the research identifies that effective AI integration depends on coordinated national strategies, teacher training and empowerment, inclusive digital infrastructure, and ethically designed tools. These findings suggest that without effective governance and human-centered implementation, AI risks reinforcing existing disparities rather than promoting equitable and sustainable education. Accordingly, this paper examines the impact of AI in education, exploring its applications, ethical considerations, and global policy frameworks.

Security Vulnerabilities and Resilience Strategies in Healthcare IoT Systems: A Comprehensive Review

Tue, 11 Aug 2026 00:00:00 GMT

Internet of Things (IoT) technologies in the healthcare industry, also known as the Internet of Medical Things (IoMT), have proven to greatly improve patient monitoring, diagnostics, and clinical decision-making. The increasing prevalence of resource-challenged medical devices, wireless connectivity, and cloud services, however, has brought new risks around security and privacy concerns that can now directly impact patient safety and data integrity. In this paper, a thorough study of 41 peer-reviewed research papers from January 2018 through May 2025 revealed the current state of security vulnerabilities and resilience strategies in healthcare IoT systems. It provides a comprehensive analysis of security threats at the device, network, and application levels such as unauthorized access, malware and ransomware, data breaches, and denial-of-service attacks delivered in a systematic manner. This contrasts with existing surveys, which consider single security mechanisms and improve upon various multi-layered security means such as AI-enabled anomaly detection, blockchain-based authentication and auditability, low-compute cryptographic techniques, and privacy-preserving methods such as federated learning. The outcomes also show that although emerging technologies add a great deal of security and trust capabilities, issues on scalability, interoperability, deployment, and regulations are not yet fully addressed. This review highlights important knowledge gaps and offers structured knowledge and future directions for research to address the design of secure, resilient, and practically deployable IoMT architectures for real-world healthcare environments.

Benchmarking Pre-Trained Feature Extractors: A Comparative Study Across Deep Learning Tasks

Tue, 11 Aug 2026 00:00:00 GMT

The rapid growth of artificial intelligence and deep learning has revolutionized diverse domains, including computer vision where transfer learning through pre-trained models has become a fundamental technique for enhancing performance of models and reducing computational costs. Although numerous pre-trained deep learning models have been widely adopted, their effectiveness varies significantly across application domains and datasets. The lack of comprehensive comparative studies makes it challenging for researchers to identify the most suitable models for achieving optimal performance in specific tasks. This study systematically compares seven pre-trained feature extractors across three architectural families, convolutional neural networks (CNNs), Vision Transformers (ViTs), and self-supervised models to provide practical guidance on model selection for downstream deep learning tasks. These models were evaluated on five benchmark datasets. Features were extracted with frozen weights and evaluated using linear probing, k-nearest neighbor retrieval, and 5-shot classification. ConvNeXt-B achieved the highest mean linear probe accuracy (90.40%), while DINOv2-S produced the best feature geometry for retrieval tasks (87.64%). CLIP-ViT-B/32 demonstrated the strongest cross-domain transfer, leading on texture recognition and satellite imagery few-shot classification. Older CNN architectures lagged significantly behind modern models by approximately 7%. Overall, this study simultaneously evaluates multiple feature extractors across multiple visual domains and evaluation protocols.

Effects of Farm Formulated Hay Feeds on Growth Performance of Heifer

Sat, 31 Jan 2026 00:00:00 GMT

Optimal nutrition during early life is critical for maximum growth and development of heifer calves. Farm-made hay feeds are widely used in small- to medium-scale farming due to their low cost and availability; however, their effects on calf growth performance are not well established. This study evaluated the impact of different hay types on growth performance in heifer calves. Twenty healthy 12-week-old heifer calves were randomly assigned to four groups (n = 5 per group). Groups A, B, and C were fed Kikuyu, Paspalum, and Desmodium hay, respectively, along with a commercial calf starter, while the control group received a commercial Total Mixed Ration (TMR). The trial lasted four weeks, with weekly body weight measurements. In week 2, weight gain was significantly higher in the TMR group than in the Desmodium hay group (p < 0.05). During weeks 3 and 4, the TMR group gained significantly more than all hay-fed groups. TMR-fed calves recorded the highest Average Daily Gain (ADG) with a consistent upward trend, whereas hay-fed calves showed a declining trend. Although TMR enhanced growth, body weights remained below optimal targets, indicating the need for improved diet formulations and optimized hay-to-starter ratios to achieve desired growth benchmarks in heifer calves.

Physicochemical and Sensory Responses of Cobb 500 Broiler Meat to Varying Pre-Slaughter Holding Periods

Sat, 31 Jan 2026 00:00:00 GMT

This study investigated the impact of three holding durations 1 hour, 2 hours, and 3 hours on the physicochemical, textural, and sensory properties of Cobb 500 broiler meat, using birds of uniform age and rearing conditions. Quality parameters assessed pH, moisture content, cooking loss, drip loss, crude protein, ash content, and colour values (L*, a*, b*), alongside textural attributes; adhesiveness, chewiness, springiness, gumminess, and cohesiveness. Statistical analyses were conducted at a significance level of p < 0.05. The results demonstrated that 3 hours holding period enhanced moisture content and meat brightness while reducing drip loss, though it also increased chewiness and gumminess. In contrast, 2 hours holding period stated higher protein retention, reduced cooking loss, and acceptable tenderness. 1 hour holding group consistently exhibited the least favourable outcomes across measured attributes. Sensory evaluation, analysed using the Friedman test (p < 0.05), confirmed that the 2 hours treatment achieved the highest scores for tenderness, flavour, texture, and overall acceptability. Colour and juiciness did not differ significantly among treatments; 2 hours group consistently outperformed the others. Findings indicate that a 2 hours pre-slaughter holding period represents the optimal strategy for improving both physicochemical and sensory quality of Cobb 500 broiler meat.

Evaluation of Volatile Compounds in Rice Bugs (Leptocorisa acuta) and Assessing their Behavioral Responses to Different Sri Lankan Rice Varieties

Sat, 31 Jan 2026 00:00:00 GMT

Rice bug (Leptocorisa acuta) is a major pest responsible for significant yield losses in rice. Although conventional chemical insecticides are effective, they pose environmental and health hazards, highlighting need for exploring alternatives such as semiochemicals and bio-insecticides. In Sri Lanka, no prior studies to identify the Volatile Organic Compounds (VOCs) associated with rice bugs. We aimed to characterize VOCs emitted by L. acuta and its behavioral responses over eight Sri Lankan rice varieties (BG 251, BG 257, BG 377, AT 313, BG 360, BG 366, AT 311, AT 362). VOCs were collected from male and female rice bugs (20 individuals per sex, 3 replicates) using dynamic headspace sampling and analyzed using Gas Chromatography-Mass Spectrometry (GC-MS). D-limonene identified as the dominant in females, while 2-octenal was in males. Behavioral assays using a four-choice olfactometer revealed varietal preferences: BG 377 was the most attractive (5.0 ± 0.89, 6.88 ± 2.33, 6.88 ± 1.67), while BG 360 was the least preferred (3.5 ± 0.33, 3.25 ± 0.78, 4.25 ± 0.73). However, statistical significance was only in one instance (p = 0.046), where others non-significant (p = 0.264, p = 0.552). These findings can be used for future development of semiochemical-based pest management.

Evaluating the Larvicidal Efficiency of Selected Botanical Plant Extracts against Dengue Vector Aedes sp.

Sat, 31 Jan 2026 00:00:00 GMT

Growing numbers of mosquito-borne diseases worldwide, especially dengue fever, demand sustainable and environmentally friendly control measures. This study compared the larvicidal effects of five medicinal plant extracts-Nicotiana tabacum (Tobacco), Acorus calamus (Sweet flag), Pongamia pinnata (Pungam), Calotropis sp., and Aristolochia bracteolata (Worm killer)—against Aedes sp. larvae. Methanolic extracts were tested at concentrations ranging from 25 to 500 ppm, and larval mortality was recorded after 24 h of exposure. Acorus calamus exhibited the highest larvicidal efficacy, with 98.00 ± 2.45% mortality at 500 ppm and the lowest lethal concentration values (LC50 = 58.07 ± 0.80 ppm; LC90 = 557.82 ± 0.75 ppm; LC99 = 1103.22 ± 0.75 ppm). Calotropis sp. showed moderate activity with an LC50 of 97.27 ± 0.75 ppm. One-way ANOVA revealed significant differences in larval mortality among plant extracts (p < 0.05), and Tukey post hoc analysis indicated that Acorus calamus differed significantly from Pongamia pinnata and Aristolochia bracteolata (p < 0.05). Regression analysis showed a significant dose–response relationship for all extracts, with the highest regression coefficient for Acorus calamus. Phytochemical screening identified phenols, terpenoids, and alkaloids as major bioactive components. Overall, Acorus calamus demonstrates strong potential as a natural larvicide for mosquito control programs.

A Comparative Evaluation of the Nutritional, Functional, and Sensory Properties of Paneer Enriched with Selected Green Leaves Additives

Sat, 31 Jan 2026 00:00:00 GMT

Paneer, is a fresh dairy product, is valued for its high protein content and versatility. It lacks essential micronutrients, necessitating strategies to incorporating nutrient-rich leafy greens to enhance its nutritional and functional properties. This study evaluates the effects of incorporating moringa (Moringa oleifera) leaf powder (T2), sesbania (Sesbania grandiflora) leaf powder (T3), and spinach (Spinacia oleracea) leaf puree (T4) into paneer, compared to a control (T1) sample. Proximate analysis indicated higher levels of protein, ash, and carbohydrates in the enriched samples. Spinach-enriched paneer exhibited the highest protein content (26.46%), followed by sesbania (25.88%), and moringa (25.62%). And moringa-enriched paneer had the highest ash content (1.97%). The antioxidant activity was significantly enhanced in enriched samples, with moringa-enriched paneer exhibiting the highest (71.30%), followed by sesbania (62.54%), and spinach (51.66%). Significant differences were observed in textural properties, the highest hardness was recorded in moringa-enriched paneer (8.88N), while sesbania-enriched paneer exhibited the highest chewiness (0.02J) and gumminess (6.57N). Total plate count revealed no significant differences (p > 0.05) among four treatments. Overall, the incorporation enhanced the nutritional and antioxidant potential of paneer without compromising its microbial safety or sensory attributes, demonstrating their potential as functional fortifying agents in dairy-based foods.

Integrated Approaches for Sustainable Management of Fungal Diseases in Chamomile Plants in Hill Country; Sri Lanka

Sat, 31 Jan 2026 00:00:00 GMT

Matricaria chamomilla (chamomile) is an important medicinal plant widely used in medicine, cosmetics, and the food industry. Fungal diseases pose a major challenge to chamomile cultivation, particularly in the hilly regions of Sri Lanka. This study focused on management of powdery mildew (Erysiphe cichoracearum), a destructive fungal disease affecting chamomile growth. A field experiment was conducted using a bio-fungicide, a vinegar solution, a chemical fungicide (Mancozeb), and an untreated control. The experiment followed a Randomized Complete Block Design (RCBD), with 20 plants per block and a total of 80 plants across four treatments. Plant height, average leaf size, and disease severity were assessed. The bio-fungicide treatment produced the highest plant growth (8.02 ± 0.21 cm) and improved overall plant health. The vinegar treatment also showed promising results (7.45 ± 0.18 cm), outperforming both Mancozeb and the untreated control. Mancozeb effectively reduced disease severity (7.27 ± 0.14 cm) but resulted in comparatively lower plant growth. The untreated plants showed severe disease symptoms, yellowing leaves, and stunted growth. The results highlight the importance of integrated disease management strategies that combine biological and eco-friendly approaches to enhance chamomile productivity while minimizing chemical fungicide use and promoting sustainable cultivation practices under local conditions.

Identifying Employee Emotions Based on Facial Expression Analysis in IT Sector using Ensemble Learning Techniques: A Systematic Literature Review

Sat, 31 Jan 2026 00:00:00 GMT

Facial expression analysis is an important tool for determining employee emotions, especially in the IT sector, where knowing emotional states can have significant effects on productivity and well-being. Traditional techniques for emotion detection are frequently limited by their dependence on simple recognition models that may not fully represent the complexities of human emotions. However, recent research has investigated the possibilities of ensemble learning techniques to improve the accuracy and dependability of these systems. This systematic literature review examines the various approaches used in emotion recognition from 2011 to 2023, with a focus on how ensemble learning might increase the efficiency of facial expression analysis. An initial selection of 33 studies was taken from six internet databases, with 7 eventually being included in the final review. The findings show that integrating several analytical methodologies into models produces promising results, leading the path for more robust emotion identification systems in the workplace.

Evaluation of Pathogen Characteristics and Management Strategies of Tomato Bacterial Stem Rot under Protected Cultivation in Sri Lanka

Sat, 31 Jan 2026 00:00:00 GMT

Tomato bacterial stem rot, caused by Pectobacterium carotovorum subsp. carotovorum, is a major constraint in protected tomato cultivation in Sri Lanka, particularly under humid polytunnel conditions. This study aimed to characterize the causal pathogen, evaluate resistance among tomato germplasm, and assess the efficacy of selected copper-based compounds for disease management. Five bacterial isolates were obtained from stem rot-infected tomato plants collected from polytunnels in the Kandy and Matale districts and cultured on Potato Dextrose Peptone Agar (PDPA). All isolates were identified as P. carotovorum subsp. carotovorum based on colony morphology, Gram reaction, catalase activity, pathogenicity on tomato plants, hypersensitivity reactions on tobacco leaves, and virulence on carrot slices. Greenhouse screening of 25 tomato germplasm lines revealed disease incidence ranging from 10-65%, with no fully resistant lines identified; however, AVTO 2117 showed the lowest disease incidence (10%) and was classified as moderately resistant. An in vitro poison food assay demonstrated complete inhibition of bacterial growth by copper hydroxide. These findings suggest that moderately resistant germplasm and copper hydroxide have potential as components of integrated management strategies for tomato bacterial stem rot; however, field-based validation is required to confirm their effectiveness under cultivation conditions.

Exploiting VSFTPD Backdoor Command Execution in Metasploitable 2

Wed, 31 Dec 2025 00:00:00 GMT

This paper explains cybersecurity vulnerabilities in legacy systems, focusing on a critical backdoor in Very Secure FTP Daemon (VSFTPD) version 2.3.4. Using Metasploitable 2 as a controlled test environment, this research demonstrates systematic penetration testing with Kali Linux and Metasploit Framework to exploit the VSFTPD backdoor vulnerability. This experiment achieved complete administrative control of the target system successfully, exposing the severe risks of unpatched vulnerabilities in production environments. Key findings consist detailed attack vectors, exploitation techniques, and system compromise methods that malicious actors could utilize. Furthermore, emphasizes important defensive measures including regular security patching, configuration hardening, and active vulnerability management. This paper answers ethical considerations in penetration testing and advocating for responsible disclosure policies including authorized security assessments. This experimental study contributes to understanding legacy system weaknesses and provides real world practical solutions for organizations managing similar infrastructure. This experiment exhibited methods attackers use to conceal their identity. This paper identified the importance of repetitive security awareness training, regular vulnerability assessments, and importance of defence-in-depth strategies to prevent exploitation of known backdoor vulnerabilities in network services. These results provide valuable insights for organizations discovering to strengthen their cybersecurity posture against legacy system threats.

Solar Versus Sun Drying of Mixed-Species Fish Waste: Effects on the Nutrient Composition of Fish Meal

Wed, 31 Dec 2025 00:00:00 GMT

This study investigated the nutrient composition of fish meal produced from locally collected fish waste (a mixture of species) using two methods: a solar drier and open sun drying. The resulting meals were compared with Imported Fish Meal (IFM) and a locally produced Commercial Plant Fish Meal (CPFM). Moisture and protein content differed significantly (p<0.05) among the solar-drier, sun-dried, IFM, and CPFM samples. Ash content was significantly different (p<0.05) between CPFM and IFM but did not differ significantly between the solar-drier and open sun-dried samples. Fat content was significantly higher (p<0.05) in CPFM (12.82%) than in the open sun-dried, solar-drier, and IFM samples. Energy content also varied significantly among all sample types. Overall, fish meal produced using the solar dryer was superior to the open sun-dried meal in terms of protein, fat, ash, moisture, and energy content. Comparison with IFM and CPFM indicated that solar-dried fish meal is an acceptable protein source for poultry feed formulation. However, its inclusion in feed requires caution due to its higher ash content, which may affect the protein efficiency ratio in chicks.

Real-Time Sri Lankan Static Sign Language System using EfficientNet-B0

Wed, 31 Dec 2025 00:00:00 GMT

Sign Language serves as an important medium of communication for the deaf and hard of hearing community. However, there is a lack of understanding among the public regarding the various signs used. Although around 300,000 Sinhala-speaking individuals are either deaf or hard of hearing, dedicated applications for Sri Lankan Sign Language detection are quite limited. In contrast, various sign language detection systems for other popular sign languages, such as American Sign Language (ASL) and British Sign Language (BSL), exist in abundance. This study aims to design and develop a lightweight mobile application, backed by the Efficient-B0 model, for Sri Lankan Static Sign Language detection. This application can detect five popular words using nine static signs and can be scaled up in the future. The developed application was found to be lightweight, efficient, and usable among the target users, achieving an accuracy of over 90% for real-time sign detection.

Effects of Protein Extracted from CO-3 Grass on Weight Gain of Broiler Chicken

Wed, 31 Dec 2025 00:00:00 GMT

This study evaluated the effects of protein extracted from CO-3 grass on broiler growth in addition to comparing the protein extract from Super Napier, Tridax daisy, and Guinea grass types. Protein extraction was conducted using a green biorefinery approach. Kjeldahl method was used to analyze crude protein content in the extract. A feeding trial was conducted where the control group fed with commercial feed and the treatment group additionally supplemented with 10% CO-3 protein extract dissolved in the drinking water to ensure uniform intake. Each group consisted of 4 replications, with 5 chicks per replicate, monitored over 30 days. In the first week, both groups were fed equally. In the treatment group, the daily commercial feed quantity was reduced by 10%, and this reduction was replaced with 10% CO-3 green protein extract to maintain the total daily feed intake for rest of the three weeks. Weight gain was assessed weekly, and the data was analyzed using an independent t-test with SPSS. It was found that CO-3 grass had the highest crude protein content (22.20 ± 1.34%) compared to other tested grasses. Further, the weekly weight gain from the second week onwards not significantly different between control and treatment groups. These findings suggest that protein extract from CO-3 grass may have potential as a supplemental protein source for broiler chickens.