iPURSE 2026

Permanent URI for this collectionhttps://ir.lib.pdn.ac.lk/handle/20.500.14444/8203

Browse

Now showing 1 - 20 of 25

An integrated computer vision framework for structural defect identification, quantification and multi-parameter severity assessment in reinforced concrete beams
(University of Peradeniya, Sri Lanka, 2026-08-20) Bandara, U. L. N. T.; Ariyaratne, H. M. P. B.; Jayasinghe, J. A. S. C.; Jayasinghe, U.
Current computer vision applications in Structural Health Monitoring (SHM) predominantly focus on binary crack detection, offering limited support for engineering condition assessment, prioritization of repairs, and maintenance decision-making. Existing studies have demonstrated reliable crack quantification; however, multi-parameter severity assessment frameworks that convert measured crack characteristics into actionable condition ratings remain comparatively underexplored. To address this gap, this study developed a comprehensive image-based framework integrating automated structural defect identification and precise geometric quantification directly with a robust severity assessment model tailored for Reinforced Concrete (RC) beams. A benchmark dataset of surface cracks exclusively in RC beams was constructed using images captured under realistic field and laboratory conditions, intentionally including varying illumination, texture, shadows, staining, and surface deterioration commonly encountered during inspections. A total of 300 high-resolution images were collected, with 100 images manually annotated for semantic segmentation. The dataset was divided into training, validation, and testing subsets of 70%, 20%, and 10%, respectively. The finalized dataset includes precise crack masks and image-derived parameters such as crack width, length, orientation, location, density, and branching characteristics. Semantic segmentation and advanced image-processing techniques were utilized to automatically extract these physical parameters, prioritizing critical serviceability thresholds and the spatial distribution of cracks. Additionally, image preprocessing procedures, including contrast enhancement and noise reduction, were successfully evaluated to improve visibility under challenging conditions. The quantified crack characteristics were subsequently integrated into a novel rule-based severity assessment framework informed by established engineering standards and structural judgement. By evaluating multiple interrelated parameters rather than relying on crack width alone, the framework successfully executed a severity assessment that classified the affected RC beams into Minor, Moderate, and Critical severity levels. Results demonstrated that integrating semantic segmentation with multi-parameter analysis successfully extracted geometric features capable of driving these accurate, standardized severity classifications. Model outputs were thoroughly benchmarked against manual measurements and engineer-assigned condition ratings, confirming high classification reliability. Ultimately, this framework successfully converted visual defect observations into actionable engineering ratings, significantly reducing reliance on subjective manual interpretation while providing a scalable methodology for rapid inspection, maintenance prioritization, and long-term monitoring of ageing RC beam infrastructure.
Accuracy comparison of machine learning and statistical methods for daily rainfall estimation in the Kelani River Basin, Sri Lanka
(University of Peradeniya, Sri Lanka, 2026-08-20) Kamburugamuwa, S. W.; De Silva, M. M. G. T.; Herath, H. M. D. K.; Neluwala, N. G. P. B.
Accurate estimation of daily rainfall at local scales remains a major challenge in hydrology due to the highly nonlinear and spatially variable nature of precipitation. For flood forecasting and water resource management, accurate rainfall prediction is essential. This study evaluates the performance of machine learning and conventional techniques for estimating daily rainfall at the Hanwella station, located within the Kelani River Basin, based on rainfall data from four neighbouring stations. An Extreme Gradient Boosting (XGBoost) model and a Long Short-Term Memory (LSTM) model were developed using multi-station rainfall data from 2020 to 2025, while spatial interpolation techniques such as Inverse Distance Weighting (IDW) and the Normal Ratio method were used as reference techniques. The LSTM model was trained using 3-day temporal sequences with various loss functions (Mean Absolute Error (MAE), Mean Squared Error (MSE), Huber loss, Mean Squared Logarithmic Error (MSLE), and Log-Cosh loss). The performance of all four models was assessed using Root Mean Square Error (RMSE), MAE, and Nash–Sutcliffe Efficiency (NSE). These evaluation metrics were calculated separately for the validation and testing datasets to provide a comprehensive comparison of model performance. Among the evaluated methods, the IDW approach (p = 2) exhibits the highest performance, achieving an RMSE of 18.32 mm/day and an NSE of 0.63, indicating a moderate level of predictive accuracy, whereas the remaining methods demonstrate comparatively lower predictive accuracy. XGBoost demonstrated competitive performance on the structured tabular dataset, followed by the Normal Ratio method, while LSTM showed the lowest overall accuracy. This behaviour may be attributed to the dominance of spatial relationships over temporal dynamics, thereby limiting the effectiveness of sequence-based learning. These outcomes imply that model configuration, feature representation, and data properties significantly impact model performance. The LSTM model has difficulty predicting peak rainfall events, even if it was able to capture overall rainfall variability. XGBoost shows superior performance on the structured dataset compared to LSTM, but both methods can be further improved. This study helps to make more accurate flood predictions and water resource management by offering useful insights for enhancing rainfall estimation techniques.
Accuracy comparison of lightweight vision-language models for multimodal fake news detection
(University of Peradeniya, Sri Lanka, 2026-08-20) Ekanayake, E. M. C. P.; Herath, D.
Fake news shared on digital platforms increasingly combines textual and visual content to spread misleading information. The Indian Fake News Dataset (IFND) is a multimodal benchmark for fake news detection in the Indian news domain. The use of lightweight vision-language models for multimodal fake news detection on the IFND dataset has not yet been explored. We focus on lightweight vision-language models because fake news detection on social media platforms requires models that can process large amounts of multimodal content efficiently and with reasonable computing cost. In this work, we conduct a comparative analysis of lightweight vision-language models for binary fake news detection on IFND. Each sample is represented as a caption-image pair, and the model predicts whether the multimodal content is true or fake. The study is conducted on a subset of IFND containing 25,224 samples. Our experiments compare the performance of six lightweight vision-language models on this benchmark. Among the evaluated models, InstructBLIP achieved the best test accuracy of 72%, while Qwen-VL showed the lowest test accuracy of 33%. These results show that lightweight vision-language models can be applied for multimodal fake news detection, although their performance differs considerably across model architectures. Overall, this work presents a benchmark of lightweight vision-language models on IFND and highlights their potential for practical multimodal misinformation detection.
A wireless soft PLC architecture for scalable and cost-efficient intelligent automation platform for precision agriculture
(University of Peradeniya, Sri Lanka, 2026-08-20) Siriwardena, S. J.; Bandara, A. S.; Ekanayake, J. B.; Samaranayake, B. G. L. T.
Intelligent automation in precision agriculture demands scalable, adaptive, and cost-effective control architectures capable of reliable operation across distributed environments. Conventional Programmable Logic Controllers (PLCs), while deterministic and robust, are limited by high cost, hardware dependency, and poor scalability in geographically dispersed applications. This work proposes a software-defined Soft Programmable Logic Controller (Soft-PLC) architecture that virtualizes traditional PLC functionality while maintaining deterministic execution and hierarchical organization. A unified cyber-physical framework is introduced using a single long-range LoRa communication network to support sensing, closed-loop control, and actuation over distances up to 1 km. Wireless sensor nodes, each powered by a dedicated solar panel with a backup battery pack, measure nitrogen, phosphorus, potassium, pH, temperature, humidity, and electrical conductivity. Data are transmitted to a centralized control hub at intervals ranging from 30 seconds to 5 minutes, depending on agronomic requirements. The unified bidirectional LoRa infrastructure replaces the multiple communication gateways and multi-network infrastructure of conventional PLC-based precision agriculture systems, reducing system complexity and communication hardware by 30– 50%. Control logic is executed on a Raspberry Pi-based hub using a deterministic scan-cycle paradigm (input acquisition, logic execution, output update) aligned with IEC 61131-3 standards, achieving local controller scan times of 40–80 ms. End-to-end communication performance is evaluated separately through LoRa network latency and packet loss. The system follows ISA-95 layered design principles to ensure structured interaction between field devices, control processes, and supervisory systems, supporting integration with existing industrial automation frameworks. The key contribution is the integration of Soft-PLC virtualization, unified LoRa communication, and ISA-95 architecture into a scalable, hardware-independent automation platform for precision agriculture. Field validation over 72 hours in an open agriculture field with node–hub distances from 200 m to 1 km demonstrates the feasibility and stable short-term operation of the proposed architecture, with an average communication latency of 180 ms (range: 150–250 ms) and a packet loss rate of 2.3% (below 5% across all nodes). The system achieves a 40–50% cost reduction compared to conventional PLC-based precision agriculture setups while maintaining comparable control performance, providing a scalable automation and communication platform capable of supporting future AI-driven precision agriculture applications.
A unified comparative analysis of attention mechanisms for pneumonia detection in chest X-ray images
(University of Peradeniya, Sri Lanka, 2026-08-20) Selvarasa, T.; Ramanan, A.
Lower respiratory tract infections continue to represent a significant global health burden, particularly among children in low-resource settings. Limited access to expert radiological interpretation further increases the need for reliable automated diagnostic systems using chest X-ray (CXR) imaging. This study presents a systematic and controlled comparative analysis of three widely used attention mechanisms: Squeeze-and-Excitation (SE), Spatial Attention (SA), and Convolutional Block Attention Module (CBAM) within a unified deep learning framework for pediatric pneumonia detection. Unlike existing studies, many approaches report results using inconsistent datasets, preprocessing methods, and training strategies, making fair comparison difficult. To address this limitation, the present study ensures fairness and reproducibility by integrating each attention mechanism into the same ResNet50 backbone architecture and training all models under identical experimental conditions. The study utilises a publicly available Chest X-ray dataset (https://tinyurl.com/kaggle-CXR) comprising 5,856 pediatric CXR images, categorised into pneumonia and normal classes, using the predefined patient-wise train, validation, and test splits to prevent data leakage and ensure fair evaluation. Consistent preprocessing and data augmentation techniques are applied to improve model generalisation. Each attention module is incorporated across all hierarchical stages of the network, enabling progressive refinement of feature representations from low-level visual patterns to high-level disease-specific characteristics. Model evaluation is conducted using accuracy, precision, recall, and F1-score, along with Grad-CAM visualisations to assess interpretability and highlight clinically relevant lung regions. An ablation study is conducted to analyse the contribution of attention placement across network stages and evaluate the impact of channel and spatial attention components. The experimental findings indicate that all attention-based models outperform the ResNet50 models, confirming the effectiveness of attention mechanisms in improving classification performance. ResNet50 achieved an accuracy of 91.03%, while ResNet50+SE, ResNet50+SA, and ResNet50+CBAM achieved 91.99%, 92.31%, and 92.95%, respectively. ResNet50+CBAM improved accuracy by 1.92 percentage points over ResNet50. Test results confirm this gain stems specifically from combined channel-spatial refinement across all stages, not either component alone. These findings offer the first controlled evidence that CBAM's advantage holds specifically for pediatric pneumonia detection.
A regenerative braking system for an open wheeled single seater rear wheel drive formula-type race car
(University of Peradeniya, Sri Lanka, 2026-08-20) Pilapitiya, P. K. W. R. M. T. R. B.; Wedisinghe, W. A. T. A.; Thajudeeen, M. T. M. M.; Weerakoon, W. M. A. A. B.; Ampitithanne, A. T. G. S. K.; Illankoon, H. M. D.; Herath, H. M. A. D. K.
Braking systems in high-performance race cars must deliver consistent stopping power, precise modulation, and high thermal resistance under extreme operating conditions. Modern motorsport increasingly emphasizes energy efficiency and sustainability, driving the adoption of regenerative braking technologies. Regenerative braking converts kinetic energy, generally lost as heat during deceleration, into electrical energy that can be stored and reused, improving vehicle efficiency. Consequently, regenerative braking has become vital in motorsport for improving efficiency and reducing energy losses. This project presents the design, development, integration, and validation of a combined conventional hydraulic and regenerative braking system for a single-seater rear-wheel- drive race car named Diablo XI, developed for Formula motor sports competitions in Sri Lanka. The project focused on advanced engineering integration, improved efficiency, and cost-effective safety enhancements. A novel belt-driven, generator-based regenerative braking mechanism is proposed, enabling energy recovery in a cost-effective manner suitable for internal combustion engine-powered race vehicles. SolidWorks, the computer-aided design and simulation tool, was used to model the system prior to physical implementation. The project used conventional braking system consists of ventilated disc brakes on all four wheels, a vacuum-assisted brake booster, a dual-line master cylinder, semi-metallic brake pads, production-based calipers, steel brake lines, and DOT 3 brake fluid. The regenerative braking system utilizes a modified alternator operating as a generator, driven via a belt connected to the engine crankshaft. Generator engagement is controlled through an overrunning clutch activated during braking events. The recovered electrical energy is stored in a 12V lithium-ion battery pack equipped with a battery management system (BMS). The hydraulic braking system was validated through static inspections and dynamic testing at speeds of up to 100 km/h. Braking performance under repeated racing conditions was evaluated, and the test results confirmed reliable braking performance, consistent generator engagement during braking events, and successful battery charging while maintaining vehicle safety, stability, and handling characteristics. Future work will focus on comprehensive track testing and further optimization of the regenerative braking strategy to maximize energy recovery and braking performance under competitive racing conditions. Overall, this study demonstrates the feasibility of implementing a low-cost regenerative braking system for student-level and grassroots motorsport applications.
A multi-agent AI framework for real-time flood disaster response using satellite imagery and geospatial analysis
(University of Peradeniya, Sri Lanka, 2026-08-20) Abeywardena, M. B.; Dilshan, R. W. V. I.; Tharaka, A. L. S.; Iddamalgoda, I. H. R. P.
Flood disasters demand rapid situational awareness spanning hazard detection, infrastructure disruption, evacuation planning, and community coordination, yet conventional GIS workflows depend on manual tools and specialist intervention, delaying critical decisions for first responders in resource-constrained emergency environments. This study aimed to design, implement, and evaluate an autonomous multi-agent AI framework that converts satellite imagery and geospatial data into actionable, near-real-time flood-response intelligence usable by non-GIS-expert responders, integrating vision-language modelling, multi-agent orchestration, and geospatial analytics into a single decision-support pipeline. The system employed four specialized agents orchestrated through CrewAI with a Llama 3.2 reasoning backbone. A Qwen2-VL-7B vision-language model was finetuned using Low-Rank Adaptation on an AMD Instinct MI300X GPU with the Cyclone Ditwah 2025 Sentinel-1 SAR dataset (50,000 samples; 45,000/5,000 train/validation split) for damage-zone classification and GeoJSON generation, while live precipitation from the Open-Meteo API and OpenStreetMap road networks processed via GeoPandas and NetworkX supported blocked-road detection, shelter identification, and hazard-avoidant routing. Evaluation used 5,000 validation scenes and 50 routing scenarios. The fine-tuned model achieved 83.70% mAP@0.5 (82.51% precision, 86.77% recall) for damage-zone detection, generating valid GeoJSON outputs in under five seconds per image; the pipeline produced complete situational reports in under sixty seconds, and the routing module generated topologically valid evacuation paths in 41 of 50 cases. Supporting modules delivered early warning alerts, seven-day forecasting, geotagged incident reporting, missing-person tracking, SOS dispatch, and offline HTML map export. These results show that combining fine-tuned vision-language models, multi-agent orchestration, and geospatial analytics can consolidate fragmented flood-response workflows into a unified, near-real-time decision-support process suitable for resource-constrained emergency environments.
A comparative analysis of the energy demands and runtime of Large Language Models when deployed on user-side devices
(University of Peradeniya, Sri Lanka, 2026-08-20) Ahamed, H. U.; Aturusinghe, A. L. A. D. T. T.; Dewmini, D. M. K. N.; Jayarathne, M. G. S. K.
Large Language Models (LLMs) are statistical algorithms trained on massive amounts of data to predict the most appropriate response to a user prompt. LLMs requires a substantial supply of resources to function and despite the increase in the number of LLMs locally deployed, studies about LLM energy consumption are limited to those run at data centers, with research on locally deployed LLM energy consumption lacking. This study aims to bridge that gap by providing important information on the energy consumption and runtime of LLMs deployed on user-side devices via a comparison of the energy consumption and runtime efficiency of selected locally deployed LLMs across different prompt categories on consumer-grade hardware. For this task, a personal laptop having 16.0GB RAM and an Intel(R) UHD Graphics GPU was configured with Ollama for local LLM deployment and HWiNFO64 for hardware monitoring. To ensure the relevancy of the research, the study uses the 5 prompt categories most used in locally deployed LLMs (Factual, Reasoning, Long Generation, Code Generation and summarization), however future studies will use more novel prompts with increasing complexity. For proper evaluation we chose models developed by different companies (DeepSeek-R1(7B), Gemma 3(4B), and Phi-4 Mini (3.8B)). Using Ollama, each model was run 45 times, with 3 tests per prompt. During each test, HWiNFO64 recorded average power use and runtime, and these average values were compared across models and categories. The findings indicate that throughout the test cases, an overall average of 23.34 Watts (W) was consumed by all three LLMs at a mean runtime of 37.81 seconds. Furthermore, results reveal that DeepSeek-R1 (7B) had the greatest power demand with an average of 29.89 W, while Gemma 3(4B) recording the lowest demand for power with an average of 18.72 W needed. The results denote that there exists a relationship between model architecture, prompt category and PC system specifications that affects energy usage. By elaborating on this relationship and offering key insights, this study provides opportunities for future LLM development in consideration of user-device capabilities and to propose a reproducible LLM benchmark for consumer grade hardware to support future research.
A clamp-on magnetic field energy harvester for self-powered dynamic line rating sensor nodes
(University of Peradeniya, Sri Lanka, 2026-08-20) Nawarathne, H. M. D. D.; Bandara, A. S.; Siriwardena, S. J.; Thennakoon, T. M. T. N.; Ekanayake, J. B.
The increased integration of renewable energy resources into transmission networks has intensified power flow variability, exposing the limitations of static line rating methods that systematically underutilise conductor capacity by relying on conservative worst-case thermal assumptions. Dynamic Line Rating (DLR) offers a data-driven alternative through real-time environmental and electrical monitoring. However, its widespread deployment is constrained by the absence of viable, autonomous power solutions for sensor nodes situated in remote transmission infrastructure, where grid-connected supply is economically and logistically prohibitive and conventional battery systems impose recurring maintenance costs and environmental burdens. This study investigates a clamp-on Magnetic Field Energy Harvester (MFEH) as a fully self-sustaining power source for DLR sensor nodes, exploiting stray electromagnetic fields emanating from AC current-carrying transmission conductors, an energy source that is continuously available, unlike intermittent ambient sources such as solar or wind. The MFEH system comprises a toroidal inductive pickup coil, a rectifier, a fourswitch buck-boost (FSBB) converter for power conditioning, and an ESP32 microcontroller is used to implement an impedance matching algorithm for Maximum Power Point Tracking (MPPT), enabling dynamic optimisation of energy extraction across varying conductor current magnitudes and load conditions. Harvested surplus energy is directed to a supercapacitor buffer to sustain operation during low-current contingencies. Experimental characterisation across distribution line currents ranging from 20 A to 80 A demonstrated harvested power levels between 80mW to 2280mW, with a peak conversion efficiency of 70.89% in the FSBB converter at fixed impedance operation. These results provided sufficient regulated output power to continuously operate the DLR sensor node while maintaining a stable 8 V supply. These findings establish that infrastructureindependent, self-powered DLR deployment is technically feasible, offering a scalable pathway to cost-effective grid instrumentation. By enabling continuous real-time line rating without external power dependency, this approach directly supports reductions in transmission congestion, curtailment of renewable generation, and advances the operational intelligence of smart grid infrastructure.
Traditional sweetener to functional food: Improving the bioactive profile of kithul treacle through Tabernaemontana divaricata nectar-microbiome fermentation
(University of Peradeniya, Sri Lanka, 2026-08-20) Ravi, K.; Silva, U. B. K.; Samarakkody, A. H. A.; Perera, E. W. B. K.; Jayasuriya, T. L.; Sewwandi, P. A.; Dharmasiri, R. B. N.
Floral nectar harbours unique microbial communities and natural antimicrobial properties. Whereas kithul treacle, a traditional Sri Lankan sweetener with physicochemical properties similar to bee honey, has its natural food preservative potential largely understudied. This study investigated the biotechnological application of Tabernaemontana divaricata floral nectar microbiome to enhance the food preservative capabilities and bioactive profile of kithul treacle through microbial fermentation. To achieve this, bacteria from T. divaricata nectar were isolated and screened for sugar fermentation and antagonistic properties to prepare a compatible consortium that could ferment fructose, sucrose, and glucose. Selected strains were phenotypically characterized using Gram staining and molecularly characterized through 16S rRNA gene sequencing, BLAST analysis and phylogenetic mapping. The consortium was used to ferment kithul treacle for 21 days at room temperature. Fermented treacle was evaluated using a PEN3.5 E-Nose system for volatile organic compound (VOC) profiling and functional enhancements were evaluated using anti-microbial activity assays (disc diffusion) and anti-inflammatory assays (albumin denaturation). All steps except the initial culturing involved duplicates, while the anti-inflammatory assay involved quadruplicates. Initial culturing yielded 14 isolates, of which eight fermented all three sugars. The antagonistic assay identified a compatible three-strain consortium of Gram-negative bacilli, consisting of Atlantibacter hermannii (PZ138294), Pantoea sp. (PZ142719), and Klebsiella pneumoniae (PZ143343). Phylogenetic analysis placed them into distinct, genus-level clades supported by high bootstrap values ranging from 86% to 93%. The PCA biplot of E-Nose sensor readings showed a distinct shift in fermented treacle’s VOC profile, closely aligning it with that of bee honey. One-way ANOVA and Tukey’s HSD post hoc tests of inhibition zone diameters of the disc diffusion assay showed greater antimicrobial activity of fermented treacle (0.15 cm) than the control (0.05 cm) against Candida parapsilosis, with a statistically significant effect (p < 0.05). Its activity was comparable to bee honey (0.13 cm; p > 0.05). Furthermore, the anti-inflammatory assay revealed that the fermented treacle had an inhibition percentage of 87.57% at 100% concentration, exceeding the control (82.62%) and almost equalling bee honey (88.08%) in anti-inflammatory activity. In conclusion, enhanced anti-inflammatory and antimicrobial activities, together with altered VOC profiles, indicate that nectar-derived microbial fermentation improves the functional properties of kithul treacle.
The effect of two fungal species isolated from sporocarps of Marsilea minuta on paddy seed germination and early seedling growth
(University of Peradeniya, Sri Lanka, 2026-08-20) Pathirana K. P. A. S.; Gama-Arachchige N. S.
Marsilea minuta is a heterosporous aquatic fern commonly found in paddy field ecosystems. During dry periods, it produces drought-resistant sporocarps that can harbour fungi affecting paddy seed germination and seedling growth. This study identified fungi associated with M. minuta sporocarps and evaluated their effects on paddy seed germination and early seedling growth. Fungi were isolated from surface-sterilized and non-sterilized sporocarps using agar plate techniques and identified based on Commonwealth Mycological Institute (CMI) descriptions, followed by DNA sequence confirmation using the internal transcribed spacer (ITS1 and ITS2) regions. The effects of fungal spore suspension (50μL of 10⁵ CFU/mL per seed) on germination percentage, root length, and shoot length were assessed using five paddy varieties (BG 367, BG 479, Hangimutthan, Kalumurunga, and Suwadal) with four replicates of 100 seeds per each treatment, compared to an untreated control (seeds treated with sterile distilled water). Seeds were germinated for 14 days at 25°C under a 12 h light/12 h dark conditions. Germination data were analysed using the Kruskal–Wallis test followed by Dunn’s post hoc test, while root and shoot lengths were analysed using the Wilcoxon rank-sum test at a significance level of P<0.05. Two fungal species (Neopestalotiopsis clavispora and Mucor irregularis) were isolated from sporocarps. N. clavispora was detected both externally and internally and did not significantly affect germination but significantly enhanced seedling growth, increasing root length by approximately 1125% and shoot length by 13–24%, with the greatest response observed in Kalumurunga (P < 0.05). Mucor irregularis was restricted to the sporocarp surface and significantly reduced germination by approximately 6–13% across all paddy varieties, with the greatest reduction observed in BG 367 (P < 0.05). It also reduced root length by 2–8% and shoot length by 4–12% in BG 367, Hangimutthan, Kalumurunga, and Suwadal, although a slight growth increase (~4%) was observed in BG 479. These findings demonstrate that N. clavispora and M. irregularis associated with M. minuta sporocarps exert contrasting effects on paddy seed performance. Further, N. clavispora shows potential as a bio-stimulant to enhance seedling vigour, whereas M. irregularis poses a risk to seed germination in paddy ecosystems.
Temporal infection dynamics of Enterocytozoon hepatopenaei (EHP) and its association with shrimp growth performance in commercial Litopenaeus vannamei farms in Chilaw, Sri Lanka
(University of Peradeniya, Sri Lanka, 2026-08-20) Kotuwegedara, H. A. R. W.; Adhikaram, K. K. P. R. B.; Gunawardana, T. A.; De Silva S. H. N. P.; Anupama N. M. T.; De S. Jagoda, S. S.
Enterocytozoon hepatopenaei (EHP) is an economically significant microsporidian parasite causing hepatopancreatic microsporidiosis in Litopenaeus vannamei, resulting in retarded growth and severe production losses in the shrimp aquaculture industry. This study investigated temporal EHP infection dynamics and associations with shrimp growth performance and water quality in commercial L. vannamei farms in the Chilaw aquaculture zone, Sri Lanka. Following preliminary screening, twenty commercial shrimp farms were selected and categorized into four groups based on EHP detection patterns: EHP-undetected, pre-stocking detection, early-stage detection (14–28 days of culture [DOC]), and late-stage detection (>28 DOC). Twenty shrimp were collected biweekly from stocking to harvest and pooled for PCR-based detection. The EHP copy number was quantified using quantitative PCR (qPCR) targeting the PTP2 gene. Average body weight, body length, and selected water quality parameters were recorded concurrently. A linear mixed model analysis revealed EHP copy number was significantly influenced by detection stage, sampling timepoint, and their interaction (p < 0.001), EHP copy numbers progressively increased throughout the culture period in infected groups, with pre-stocking detection farms exhibiting the highest sustained infection load, while EHP-undetected farms maintained zero copy numbers throughout the study period. Average body weight and body length were significantly affected by detection stage, sampling timepoint, and their interactions (p < 0.001). Spearman’s correlation at the final harvest point (DOC 70) confirmed significant negative associations between EHP copy number and body weight (ρ = −0.720, p < 0.001) and body length (ρ = −0.550, p = 0.012) confirming impaired growth in infected groups. Among water quality parameters, total ammonia showed significant temporal and detection stage variation, whereas pH, salinity, differed significantly among detection stages (p < 0.001). These findings demonstrate that EHP infection stage at stocking and progressive copy number escalation are critical causes of shrimp growth outcomes, underscoring the need for rigorous pre-stocking EHP screening and early molecular detection in local commercial shrimp farming.
Synergistic enhancement of oxacillin efficacy against MRSA using green synthesized silver nanoparticles from Syzygium cumini leaf extract
(University of Peradeniya, Sri Lanka, 2026-08-20) Janarthani, V.; Gamagedara, T. P.
Methicillin-resistant Staphylococcus aureus (MRSA) is a major global health concern due to its resistance to beta-lactam antibiotics, including oxacillin. Nanoparticles have emerged as a promising alternative antimicrobial strategy, particularly when combined with existing antibiotics. This study aimed to synthesize and characterize silver nanoparticles (AgNPs) using Syzygium cumini leaf extract and evaluate their synergistic antibacterial activity with oxacillin against MRSA. Silver nanoparticles were synthesized via a green method by reacting 1 mM AgNO₃ with aqueous leaf extract (1:5 ratio, pH 8.15). Formation of AgNPs was confirmed by a colour change to brown and a characteristic surface plasmon resonance peak at 411 nm in UV-visible spectroscopy. Dynamic light scattering analysis revealed a mean hydrodynamic diameter of 155 nm and a high polydispersity index (0.92), indicating a heterogeneous nanoparticle distribution. An ATR-FTIR analysis confirmed the presence of functional groups from plant-derived biomolecules acting as reducing and stabilizing agents. Scanning electron microscopy was used to characterize the surface morphology and microstructure of AgNPs. Antimicrobial activity was evaluated using the disk diffusion method against four genetically confirmed MRSA strains and methicillin-susceptible Staphylococcus aureus (MSSA, ATCC 25923), with AgNPs alone, oxacillin alone and the AgNP-oxacillin combination alongside distilled water as a negative control. Silver nanoparticles exhibited concentration-dependent antibacterial activity against MRSA (p < 0.001, η² = 0.554). The combination of AgNPs with oxacillin produced significantly larger zones than AgNPs alone (p = 0.010), indicating a significant synergistic effect. Against MSSA, the combination demonstrated significantly greater activity than the individual treatments (p < 0.001, η² = 0.942). These findings highlight the potential of S. cumini-mediated AgNPs as antibiotic adjuvants and support further investigation into nanoparticle-based combination therapies to combat multidrug-resistant pathogens.
Supplementation of yeast fractions (Safmannan®) in broiler diets: effects on growth performance and nutrient digestibility under high stocking density
(University of Peradeniya, Sri Lanka, 2026-08-20) Vidanapathirana, V. S. A.; Jayawardana, B. C.; Samarakoon, R.; Karunaratne, N. D.; Jayawardene, L. P. I. N. P.; Singh, S.; Konthasinghe, K. H. M. N. B.; Guluwattha, J.; Gamage, W. V. G. D. L.; Weththasinghe, P.
Many intensive broiler productions rely on high stocking densities to maximize output, yet overcrowding elevates stress, compromises gut health, and reduces productivity. Yeast-derived postbiotics have shown promising effects in improving broiler performance as sustainable alternatives to antibiotic growth promoters. This study aimed to evaluate the effects of Yeast Fractions containing mannan-oligosaccharides and β-glucans (Safmannan®) supplementation on growth performance and ileal nutrient digestibility in broiler chickens reared under standard and high stocking densities. A twofactor factorial design was employed using 112 one-day-old Cobb 500 broilers randomly assigned to four treatments (n = 4 replicates per treatment): T1, control diet at standard stocking density (SSD; 11.11 birds/m²); T2, control diet at high stocking density (HSD; 15.32 birds/m²); T3, 0.05% yeast fractions supplemented diet at SSD; and T4, 0.05% yeast fractions supplemented diet at HSD. Birds were fed with experimental diets for 35 days. Growth performance parameters, carcass traits, relative organ weights, ileal nutrient digestibility were assessed. Final body weight and body weight gain were significantly increased with the yeast fractions supplementation under standard stocking density (P < 0.001). Carcass yield and breast muscle yield were significantly reduced in birds fed yeast-derived postbiotics under both stocking densities (P < 0.001). Relative weights of the heart, liver, caecum, duodenum, and jejunum were also significantly reduced in birds fed yeast-derived postbiotics under both stocking densities (P < 0.001). Overall, dietary supplementation with yeast fractions containing mannan-oligosaccharides and β-glucans improved growth performance in broilers reared under standard stocking density. Therefore, yeast fractions containing mannan-oligosaccharides and β-glucans show potential as a promising functional feed additive for improving broiler production.
Structural and functional insights into Rc allelic variation governing pericarp pigmentation in Sri Lankan traditional rice germplasm
(University of Peradeniya, Sri Lanka, 2026-08-20) Rathnayaka, W. K. K. I. K.; Gimhani, D. R.; Liyanage, D. S.; Kumarasiri, L. P. M.
The Rc gene encodes a basic helix-loop-helix (bHLH) transcription factor that regulates proanthocyanidin biosynthesis and pericarp pigmentation in rice. Allelic variation at this locus determines red, brown, and white pericarp colours. Pericarp colour was visually evaluated using the Rice Germplasm Characterization Catalogue (PGRC) in previous studies (33 pigmented and 14 nonpigmented). This study evaluated Rc allelic diversity and its structural-functional implications in 47 Sri Lankan traditional rice varieties using Nipponbare as a reference. The Rc gene sequences of 47 Sri Lankan rice varieties present in the International Rice Research Institute’s 3000 Rice Genome Project (IRRI 3K RGP) database were retrieved from Rice SNP-Seek database. Cluster analysis of the amino acid sequences (AA) identified nine distinct clusters, from which 24 representative varieties were selected based on amino acid length and colour (19-pigmented, 5-non-pigmented). Protein lengths ranged from 351 to 668 amino acids and correlated with pericarp colour. Full-length alleles (664–668 AA) were associated with 11 pigment varieties, while truncated alleles (351–473 AA) were linked to white and seven pigment pericarp varieties. ProtParam, SWISS-MODEL homology modelling, and PyMOL analyses revealed significant differences in physicochemical properties, molecular weight, aliphatic index, and secondary structures according to length. Domain (NCBI CDD) and motif (Motif Scan) analyses showed that all white pericarp varieties possessed only one domain and two motifs. Four white varieties carried the known 14-bp deletion at position 1408 in exon 6. Notably, Kuruluwee variety lacked this deletion but contained a C1353A substitution that introduced a premature stop codon. Sixteen pigmented varieties had two domains and four motifs, including the critical bHLH transcriptional activation-like domain, while light brown/striped varieties (6) typically showed three motifs. Two pigmented exceptions (Galawaka handeran, and Kula karuppan) with only one domain and two motifs were further analysed, revealing unique SNP(A897G) in Galawaka handeran that likely compensate for the missing domains to restore function. Statistical analyses confirmed a significant association between motif count and pericarp colour (Kruskal-Wallis test, p = 0.007).
Structural characterization of salinity-responsive OsSWEET15 allelic variants in Sri Lankan traditional rice germplasm
(University of Peradeniya, Sri Lanka, 2026-08-20) Nishshanka, U. J.; Gimhani, D. R.; Liyanage, D. S.; Kariyawasam, A. M.; Kumarasiri, L. P. M.; Gunawardana, U. D. G. S.
Salinity stress severely constrains rice productivity by disrupting ion homeostasis and metabolic balance. SWEET family sugar transporters regulate sucrose partitioning under abiotic stress, making them important targets for understanding natural adaptation mechanisms. This study investigated molecular variation in OsSWEET15 across 48 rice accessions to identify sequence-level signatures associated with salinity tolerance in traditional Sri Lankan indica rice varieties. Coding sequences from 48 rice varieties, including the Nipponbare reference genome and traditional Sri Lankan indica varieties, were retrieved from International Rice Research Institute (IRRI), 3,000 Rice Genomes Project. They were analysed using MEGA12 software for multiple sequence alignment and phylogenetic reconstruction. Representative varieties — salinity-susceptible Nipponbare, moderately tolerant Godawee, and salt tolerant Pachchaperumal — were selected for comparative analysis. Conserved domain analysis was performed using NCBI CDD, structural analysis using PyMOL, TMHMM was used to predict transmembrane helices, and Protter was used for graphical visualization of membrane topology. Multi-variety alignment of 319 amino acid sequences revealed five distinct substitution patterns clustering phylogenetically: R290H in Sigardis/Puttunellu/Chandina, I236T in Godawee/Pachchaperumal, P275T in Samba/Kuruthaseenati, combined I263T/P275T in Sudukarayal, and an N-terminal deletion spanning residues 1–24 in Murunga. Godawee, Pachchaperumal, Murunga, and Samba clustered with the established salt-tolerant landrace Pokkali. Notably, the I236T substitution, replacing hydrophobic isoleucine with polar threonine, occurred exclusively in tolerant varieties. The introduction of threonine may confer greater hydrogen-bonding capacity and potential phosphorylation sites, which could modulate local transporter conformation and improve OsSWEET15 function under salinity stress. This amino acid change may therefore, enhance structural flexibility at the TM7cytoplasmic interface, supporting more efficient sugar transport and osmo-protective responses in tolerant genotypes. Conserved domain analysis confirmed preservation of dual MtN3_slv domains and PQ-loop motifs across all varieties with highly significant E-values (10⁻²⁸ to 10⁻³). TMHMM predicted seven transmembrane helices, positioning residue 236 at the cytoplasmic TM7 interface. PyMOL structural alignment showed a global RMSD of 0.000 Å and a localized RMSD of 0.009 Å at residues 230–240, indicating spatially confined conformational divergence at the mutation site. The I236T substitution in OsSWEET15 represents a structurally plausible molecular marker of natural salinity adaptation with potential utility for developing salt-resilient rice varieties in affected agricultural regions.
Sensor-integrated tracking system for food waste management: A cost-effective mitigation strategy for the Sri Lankan hospitality sector
(University of Peradeniya, Sri Lanka, 2026-08-20) Jesika, D. M. N.; Somaratne, G. M.; Dharmaratne, N. S.; Sandanika, D. W. N.; Upathissa, T. P. G. K.; Samarasekera, M. A.
Food waste generation within the hospitality industry poses a critical threat to environmental sustainability, social equity, and economic stability. Precise waste measurement and monitoring are fundamental prerequisites for waste mitigation. Currently used manual waste measurement and reporting remains labor-intensive and low in accuracy. Although high-end AI waste-tracking solutions have been introduced, their adoption in developing countries like Sri Lanka is hindered by high implementation costs and a lack of customization for local food items. This study presents a low-cost, sensor-integrated tracking system engineered for large-scale hotel kitchens. The architecture of the system leverages Internet of Things (IoT) integration, utilizing load cells/weight sensors coupled with an HX711 amplifier and ESP32 microcontroller to facilitate accurate weight capture and seamless data acquisition. Weight data is synchronized via a Firebase cloud database to a dedicated Flutter-based mobile application, which serves as the primary user interface for hotel staff. The mobile application enables real-time selection of waste sources and categories while generating comprehensive waste reports, including edible vs. non-edible analysis across distinct operational areas and key waste indicators, such as edible food waste per cover and CO₂ emission from edible food waste to drive datadriven waste reduction strategies. System validation was conducted by testing in a real hotel kitchen environment. A paired t-test comparing the automated scale measurements against manual weight measurements yielded (p > 0.05), confirming the system’s accuracy by indicating the mean difference between the two systems was statistically insignificant. Furthermore, a Wilcoxon Signed-Rank test was done for staff feedback evaluation (p < 0.05), demonstrating significant improvements of the system over the manual method in terms of usability, data clarity, organization, and operational relevance. While commercial tracking systems require an enterprise investment of 90 million LKR plus recurring annual fees, this localized hardware system costs around 25,000 LKR per unit, providing a highly financially viable alternative. This localized, low-cost digital solution offers a scalable framework for enhancing sustainability within the hospitality industry.
Role of macrofauna in farm-level prevalence of Enterocytozoon hepatopenaei (EHP) in shrimp aquaculture systems in North Western Province of Sri Lanka
(University of Peradeniya, Sri Lanka, 2026-08-20) Adhikaram, K. K. P. R. B.; Kotuwegedara, H. A. R. W.; Kalupahana, A. W.; Gunawardana, T. A.; Anupama, N. M. T.; De S. Jagoda, S. S. S.
Enterocytozoon hepatopenaei (EHP) is the most significant microsporidian pathogen infecting Litopenaeus vannamei, causing hepatopancreatic microsporidiosis (HPM) disease. EHP causes significant economic losses in shrimp aquaculture worldwide. This pathogen was first reported in 2022 in the Puttalam District, North Western Province of Sri Lanka. Farm-level macrofaunal communities are considered to play a major role in disease transmission. Various macrofaunal groups, including crabs, molluscs, fish, and insects, are commonly associated with shrimp farming environments and natural aquatic ecosystems. This study aimed to investigate the association between macrofaunal communities and EHP prevalence across shrimp farms in the North Western Province of Sri Lanka. A total of 55 shrimp farms in the North Western Province were surveyed, and 158 macrofaunal samples were collected based on accessibility, comprising fish, molluscs, crabs, and aquatic insects. Species identification was conducted based on morphological characteristics. Organs from potential macrofaunal carriers were collected according to sample type, as follows: crabs (hepatopancreas, stomach, and gills); fish (gills); molluscs (gills and stomach); and insects (whole body). Organs were collected and pooled from five samples per farm. Detection of EHP was performed using nested polymerase chain reaction (PCR) targeting the spore wall protein (SWP) gene. The dominant macrofaunal species identified included Glossogobius giuris (n=32), Perisesarma bidens (n=54), Oreochromis niloticus (n=6), Cerithidea cingulata (n=18), Gerris spp. (n=12), Euphausiacea spp. (n=31), and Turritella spp. (n=5). Three species each of Arthropoda, Chordata, and Mollusca were identified based on morphological characters. All samples tested positive by nested PCR using the SWP primers. EHP was detected across multiple macrofaunal groups, with the highest prevalence observed in Glossogobius giuris (68.75%; 22/32), followed by Perisesarma bidens (59.26%; 32/54), Gerris spp. (58.33%; 7/12), and Cerithidea cingulata (50.00%; 9/18). EHP was not detected in Euphausiacea spp., Oreochromis niloticus, or Turritella spp. These findings provide strong evidence that macrofauna contribute to the environmental persistence and transmission of EHP in shrimp aquaculture systems and to the spread of disease into natural aquatic ecosystems. The results emphasize the need for enhanced biosecurity measures, including the management of macrofaunal populations within farms, to reduce the risk of disease spread at both the farm level and in natural aquatic ecosystems. This study advances the current understanding of EHP epidemiology and offers practical insights for improving disease control strategies in Sri Lanka's shrimp farming industry.
Prevalence and minimum inhibitory concentration profiling of amoxicillin-resistant bacteria in coastal waters of the Northern and Southern Provinces of Sri Lanka
(University of Peradeniya, Sri Lanka, 2026-08-20) Silva, U. B. K.; Sewmi, J. A. G. V.; Liyanage, G. Y.; Meddage, A. K. M. M. K.; Adhikari, P. A. A. S.; Sadupama, P. M.; Manage, P. M.
Antimicrobial resistance (AMR) represents a severe and escalating threat to global public health, complicating the effective treatment of infectious diseases. Wastewater discharge, agricultural runoff, and aquaculture effluents introduce antibiotic residues and antibiotic-resistant bacteria (ARB), making coastal waters particularly vulnerable to contamination. In Sri Lanka’s coastal waters, where such data is scarce, this study evaluated the prevalence and resistance profiles of bacteria resistant to amoxicillin (AMX). Water samples were collected in triplicate from eleven coastal sites: six in the Northern Province (Karaitivu, Anaikodai, Ponnalai, Dambakola Patuna, Point Pedro, and Dutch Fort-Jaffna) and five in the Southern Province (Polhena, Weligama, Koggala, Unawatuna, and Hikkaduwa). AMX-resistant bacteria were isolated using the standard pour plate method on nutrient agar supplemented with 60 µg mL⁻¹ AMX. Minimum inhibitory concentrations (MICs) were determined via agar dilution following Clinical and Laboratory Standards Institute (CLSI) guidelines. Total viable counts were higher in the southern sites (5.07×10⁴ - 6.75×10⁴ CFU mL⁻¹) than in the northern sites (1.21×10⁴ - 1.53×10⁴ CFU mL⁻¹; p<0.001), with the highest count observed at Hikkaduwa. Similarly, AMX-resistant bacteria were more prevalent in the south (2.03×10⁴ - 4.94×10⁴ CFU mL⁻¹) compared to the north (3.13×10³–1.11×10⁴ CFU mL⁻¹; p<0.001). Among 40 distinct isolates, high-level resistance (>360 µg mL⁻¹) was heavily prevalent, particularly in the Northern Province (95.83%) compared to the Southern Province (75.0%). While the remaining northern isolates (4.17%) displayed only low-level resistance at 60 – 120 µg mL⁻¹, southern isolates exhibited a broader spectrum of resistance profiles, showing MIC values of 6.25% at 60–120 µg mL⁻¹, 12.5% at 180 – 240 µg mL⁻¹, and 6.25% at 240–360 µg mL⁻¹. These findings reveal a complex geographic disparity in AMX resistance; while the southern coastal region demonstrates a higher overall prevalence of AMX-resistant bacteria at 60 µg mL⁻¹, the northern coastal region is identified as a significant hotspot for high-level resistance. To mitigate this spread, it is crucial to strengthen local antibiotic stewardship and upgrade wastewater treatment infrastructure across both provinces. Furthermore, it is highly recommended that subsequent research incorporate molecular techniques to map antibiotic resistance genes (ARGs) and expand the geographical scope to understand the ecological impact of AMR.
One Health insights into Staphylococcus aureus in live bird handling meat shops of Colombo Municipality : preliminary findings
(University of Peradeniya, Sri Lanka, 2026-08-20) Sewwandi, P. A. U.; Perera, I. A.; Gunasekara, Y. D.; Dissanayake, U. P. R. U.; Gunawardena, E. G. W.; Kottawatta, K. S. A.; Karunarathna, H. M. T. K.; Liyanapathirana, L. V. C.; Kalupahana, R. S.
Live bird handling meat shops (wet markets) are high-density human–animal–environment interfaces that facilitate zoonotic pathogen transmission, highlighting the importance of a One Health approach. However, there is limited data from Sri Lanka on pathogen prevalence in such settings. Staphylococcus aureus colonizes humans and animals and can spread through direct contact, contaminated surfaces, and respiratory secretions. Its ability to acquire antimicrobial resistance, including methicillin resistance in Staphylococcus aureus (MRSA), makes it a significant public health concern with potential entry into the food chain. Therefore, this study aimed to determine the occurrence of S. aureus in chickens, workers, and environmental samples from selected public and private wet markets in the Colombo Municipality using a One Health approach. A total of 220 samples were collected from 38 wet markets, including 132 nasal swabs from slaughtered broiler chickens, 56 nasal swabs from chicken handlers and 32 environmental dust swabs (collected from air exhaust areas measuring 10 cm × 10 cm). In addition, 32 air samples were obtained using the settle plate method, in which mannitol salt agar plates were exposed at each sampling location for 10 minutes at a height of 5 ft. Isolation and identification of S. aureus were performed using standard microbiological techniques, including growth on Mannitol salt agar, Gram staining, catalase testing, slide and tube coagulase tests, and DNase testing. Isolates positive for all identification tests were considered S. aureus. The highest isolation rate was observed in handlers, 30% (17/56; detected across 14/32 locations), followed by chicken samples, 9.8% (13/132; detected across 11/38 locations). No S. aureus isolates were recovered from environmental dust samples (0/32) or air samples (0/32). Antibiotic susceptibility and molecular characterization of the isolates are pending. The detection of S. aureus in both humans and chickens suggests possible bidirectional transmission within wet market environments. Although environmental samples were negative, close human–animal interaction may facilitate bacterial spread. These findings highlight the need for stricter animal handling and market hygiene practices, improved worker education, and routine One Health surveillance to monitor and reduce the spread of pathogenic S. aureus in wet markets.

Browse

Recent Submissions