Articles authored by UoM staff (Publish in scimago's Q1 journals)

Permanent URI for this collectionhttp://192.248.9.226/handle/123/19622

Note: These articles were published in scimago Q1 journals, which were Q1 journals at the time the article was published.

Browse

Now showing 1 - 20 of 758

item: Article-Full-text
Assessment of costs and benefits of green retrofit technologies
(Elsevier, 2023) Periyannan, E; Ramachandra, T; Geekiyanage, D
With the rising impact of greenhouse gas emissions, resource depletion, and the global interest in sustainability advancements within all sectors, construction industry practitioners are also interested in incorporating sustainable features and practices into their buildings. Nevertheless, most of the commercial buildings in Sri Lanka had been constructed during the unprecedented urbanization between 1995 and 2010, thus, before sustainable concepts became more prominent. Therefore, existing buildings in Sri Lanka is experiencing ever-increasing energy consumption, resulting in higher utility costs, with which green retrofitting has become imperative, notably in hotel buildings. This study, therefore, conducted an economic evaluation of three existing hotel buildings to establish an account of the cost implications and saving potentials of different green retrofit technologies. The data collected through document reviews and site visits were analysed using net present value and simple payback period calculations. Although number of retrofitting technologies have been incorporated in the selected buildings, more weight has been given to incorporating technologies to achieve energy efficiency and indoor environmental quality. Considering the financial viability, all the implemented green retrofits have a positive return on investment and less than ten years of payback period, except LED televisions. Amongst the implemented retrofits, biomass boilers, energy-efficient chillers, and solar PV systems have the highest energy-saving efficiency, followed by VFDs and LED lighting, while LED televisions have the lowest. The study's findings contribute to industry practitioners identifying the appropriate green retrofits based on the cost implications and savings potential and enhancing the sustainability of the built environments by reducing greenhouse gas emissions and depletion of natural resources.
item: Article-Full-text
A comparative study of the characteristics of hate speech propagators and their behaviours over Twitter social media platform
(Elsevier, 2023) Perera, S; Meedin, N; Caldera, M; Perera, I; Ahangama, S
The internet and social media have facilitated diverse communication genres, enabling widespread and rapid opinions-sharing. However, hate speech imposes a contemporary challenge on individuals and communities, given the user anonymity, freedom, and inadequate regulation. Therefore, it is imperative to identify the perpetrators responsible for spreading hate content and examine their behaviour to prevent and mitigate the negative impact. This study aimed to compare the characteristics of hate speech propagators and their behaviour with non-hate users on Twitter for the first time in Sri Lanka. The intrinsic and extrinsic profile features were extensively analyzed, employing Sinhala and English text analysis techniques. A corpus of 102882 posts from 530 hate and non-hate Twitter user profiles was selected for the study. This study investigates the unique characteristics of hate speech propagators and non-hate users by examining their profile self-presentation, conducting social network analysis, and analyzing sentiment and emotion through linguistic analysis. Hate users often refrained from expression, with infrequent account verification and geotagging. They tend to have a higher follower and following counts and more favourites, group memberships, and statuses than non-hate users. However, general Twitter user engagement with hate users was significantly low, with fewer likes, retweets, and replies. The limited involvement of normal users with hate content indicates that audiences can be effectively utilized to combat hate speech. The sentiment analysis between languages showed polarisation of negative tweets towards Sinhala, with the synergistic effect of English language users using positive sentiment to spread hate content. The novel findings shed light on the characteristics of hate users, facilitating their early detection and moderation of hate speech and aiding in developing algorithms to rank and categorize hate users using artificial intelligence. Moreover, it can be used for policy reforms, awareness programmes, and building social cohesion while combating hate speech.
item: Article-Full-text
A Secure and smart home automation system with speech recognition and power measurement capabilities
(Multidisciplinary Digital Publishing Institute, 2023) Irugalbandara, C; Naseem, AS; Perera, S; Kiruthikan, S; Logeeshan, V
The advancement in the internet of things (IoT) technologies has made it possible to control and monitor electronic devices at home with just the touch of a button. This has made people lead much more comfortable lifestyles. Elderly people and those with disabilities have especially benefited from voice-assisted home automation systems that allow them to control their devices with simple voice commands. However, the widespread use of cloud-based services in these systems, such as those offered by Google and Amazon, has made them vulnerable to cyber-attacks. To ensure the proper functioning of these systems, a stable internet connection and a secure environment free from cyber-attacks are required. However, the quality of the internet is often low in developing countries, which makes it difficult to access the services these systems offer. Additionally, the lack of localization in voice assistants prevents people from using voice-assisted home automation systems in these countries. To address these challenges, this research proposes an offline home automation system. Since the internet and cloud services are not required for an offline system, it can perform its essential functions, while ensuring protection against cyber-attacks and can provide quick responses. It offers additional features, such as power usage tracking and the optimization of linked devices.
item: Article-Full-text
Evaluating expressway traffic crash severity by using logistic regression and explainable & supervised machine learning classifiers
(Elsevier, 2023) Shashiprabha Madushani, JPS; Sandamal, RMK; Meddage, DPP; Pasindu, HR; Gomes, PIA
The number of expressway road accidents in Sri Lanka has significantly increased (by 20%) due to the expansion of the transport network and high traffic volume. It is crucial to identify the causes of these crashes for effective road safety management. However, traditional statistical methods may be insufficient due to their inherent assumptions. This study utilized explainable machine learning to investigate the factors that affect the severity of traffic crashes on expressways. The study evaluated two groups of traffic crashes: fatal or severe crashes, and other crashes that included non-severe injuries or only property damage. Five factors that contribute to crashes were analyzed: road surface condition, road alignment, location, weather condition, and lighting effect. Four machine learning models (Random Forest (RF), Decision Tree (DT), extreme gradient boosting (XGB), K-Nearest Neighbor (KNN)) were developed and compared with Logistic Regression (LR) using 223 training and 56 testing data instances. The study revealed that the machine learning algorithms provided more accurate predictions than the LR model. To explain the machine learning models, Shapley Additive Explanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME) were used. These methods revealed that all five features decreased the possibility of occurrence of fatal accidents. SHAP and LIME explanations confirmed the known interactions between factors influencing crash severity in expressway operational conditions. These explanations increase the trust of end-users and domain experts on machine learning models. Furthermore, the study concluded that using explainable machine learning methods is more effective than traditional regression analysis in evaluating safety performance. Additionally, the results of the study can be utilized to improve road safety by providing accurate explanations for decision-making processes for black-box models.
item: Article-Full-text
Real-time integration of microalgae-based bioremediation in conventional wastewater treatment plants: Current status and prospects
(Elsevier, 2023) Kankanamalage, G; Nishshanka, SH; Thevarajah, B; Nimarshana, PHV; Prajapati, SK; Ariyadasa, TU
With rising water scarcity leading to a risk of affecting 1.69 to 2.37 billion people in urban residents, the treatment and reuse of wastewater have been identified as one of the main avenues to preserve global water resources. Thus, wastewater treatment plants with capacities ranging from 8000 to 200,000 tons/day have been implemented to treat wastewater and discharge effluent with improved quality parameters. Nonetheless, the generation of 160,000–210,000 tons/year of sludge and the requirement for advanced treatment to achieve non-detectable residues are significant concerns for highly effective wastewater treatment. In this context, microalgae with the potential of effective nutrient removal from wastewater streams have been exploited in wastewater treatment at primary, secondary and tertiary treatment stages. Microalgae-based bioremediation generates valuable biomass with metabolites, namely lipids, proteins, and carbohydrates, which could be utilized in the value-added production of biofuels, biofertilizers, etc. Moreover, microalgae integrated wastewater treatment systems would substantially remove residual pollutants, nutrients, and pathogens with high removal efficiencies. Hence, the integration of microalgae into the conventional wastewater treatment process enhances the process sustainability while contributing to the concept of a circular bioeconomy. Nevertheless, limited studies are available on the potential of integrating microalgae in the conventional wastewater treatment plants for real-world applications, although several reviews are available in the literature focusing the microalgae-based wastewater treatment in a general context. Thus, the current review aims to address this gap in the literature by comprehensively assessing the prospects of integrating phycoremediation as the secondary and tertiary/advanced wastewater treatment processes, while discussing the challenges and future perspectives in the research domain.
item: Article-Full-text
Chitosan-graphene oxide dip-coated polyacrylonitrile-ethylenediamine electro spun nanofiber membrane for removal of the dye stuffs methylene blue and congo red
(Multidisciplinary Digital Publishing Institute, 2023) Pathirana, MA; Dissanayake, NSL; Wanasekara, ND; Mahltig, B; Nandasiri, GK
Textile wastewater accommodates many toxic organic contaminants that could potentially threaten the ecosystem if left untreated. Methylene blue is a toxic, non-biodegradable, cationic dye that is reportedly observed in significant amounts in the textile effluent stream as it is widely used to dye silk and cotton fabrics. Congo red is a carcinogenic anionic dye commonly used in the textile industry. This study reports an investigation of methylene blue and Congo red removal using a chitosan-graphene oxide dip-coated electrospun nanofiber membrane. The fabricated nanocomposite was characterized using Scanning Electron Microscopy (SEM), FT-IR Spectroscopy, Raman Spectroscopy, UV-vis Spectroscopy, Drop Shape Analyzer, and X-ray Diffraction. The isotherm modeling confirmed a maximum adsorptive capacity of 201 mg/g for methylene blue and 152 mg/g for Congo red, which were well fitted with a Langmuir isotherm model indicating homogenous monolayer adsorption.
item: Article-Full-text
Forest sound classification dataset: FSC22
(Multidisciplinary Digital Publishing Institute, 2023) Bandara, M; Jayasundara, R; Ariyarathne, I; Meedeniya, D; Perera, C
The study of environmental sound classification (ESC) has become popular over the years due to the intricate nature of environmental sounds and the evolution of deep learning (DL) techniques. Forest ESC is one use case of ESC, which has been widely experimented with recently to identify illegal activities inside a forest. However, at present, there is a limitation of public datasets specific to all the possible sounds in a forest environment. Most of the existing experiments have been done using generic environment sound datasets such as ESC-50, U8K, and FSD50K. Importantly, in DL-based sound classification, the lack of quality data can cause misguided information, and the predictions obtained remain questionable. Hence, there is a requirement for a well-defined benchmark forest environment sound dataset. This paper proposes FSC22, which fills the gap of a benchmark dataset for forest environmental sound classification. It includes 2025 sound clips under 27 acoustic classes, which contain possible sounds in a forest environment. We discuss the procedure of dataset preparation and validate it through different baseline sound classification models. Additionally, it provides an analysis of the new dataset compared to other available datasets. Therefore, this dataset can be used by researchers and developers who are working on forest observatory tasks.
item: Article-Full-text
Sedimentological observations and geochemical characteristics of paleo-tsunami deposits along the east coast of Sri Lanka in the Indian Ocean
(Elsevier, 2023) Ratnayake, A. S; Wijewardhana, T. D. U.; Haraguchi, T.; Goto, K.; Ratnayake, N. P; Tetsuka, H.; Yokoyama, Y.; Miyairi, Y.; Attanayake, A. M. A. N. B.
The 2004 Indian Ocean tsunami caused 230,000 fatalities and massive physical damage along the shorelines of the Indian Ocean. Holocene sedimentary archives along the coastline of Sri Lanka can provide evidence for similar past events. The objective of our current study is to investigate the sedimentological and geochemical characteristics of Sri Lankan paleo-tsunami deposits, and their chronology. Sediment samples were collected down to 10 m depth using core drilling at a sheltered site (low-lying swale depression) in Koddiyar Bay (Trincomalee) on the east coast of Sri Lanka. Visual stratigraphic observations and grain size analyses were carried out to identify lithological changes. Bulk sediments, shells, and wood fragments were used for 14C age dating. Seven sand layers were identified throughout the core in this marshy wetland. Sedimentological observations and geomorphological features show that sheltered/protected swales of the inner bay have a high potential for preservation of paleo-tsunami sediments. Three possible paleo-tsunami deposits were identified earlier than 700 years, ca. 2000 years, and ca. 2700 years. These paleo-tsunami sand layers exhibited relatively high total organic carbon (TOC) contents due to admixture of continental shelf and lagoon sediments. Analysis of paleo-tsunami records suggests that the Andaman Sea is vulnerable to both ‘low-risk and high-frequency’ and ‘high-risk and low-frequency’ type tsunamis, whereas other areas in the Indian Ocean are more prone to destructive and low-frequency type tsunamis. This study contributes to improving tsunami risk assessment and mitigation strategies along the east coast of Sri Lanka.
item: Article-Full-text
Development of a dual-chamber pyrolizer for biochar production from agriculturalWaste in Sri Lanka
(MDPI, 2023) Illankoon, W. A. M. A. N.; Milanese, C; Karunarathna, A. K; Alahakoon, A. M. Y. W; Rathnasiri, P. G; Medina-Llamas, M.; Collivignarelli, M. C.; Sorlini, S.
This study investigates the design and development of a pyrolysis reactor for batch-type biochar production from rice husks. The main objective is to develop an appropriate technology to regulate pyrolysis temperature and biomass residence time that can be easily operated under field and household conditions with minimal operational and technical requirements. The designed novel dual-chamber reactor comprises two concentrical metal cylinders and a syngas circulation system. The outer cylinder is for energy generation and the inner one is for pyrolysis. Temperature profiles, energy exchanges, syngas production, and the physicochemical characteristics of biochar were obtained to determine the performance of the reactor. Different trials were carried out to obtain different pyrolysis temperatures under constant amounts of feedstock and fuel. The temperature was monitored continuously at three predetermined reactor heights, the temperature profile varied from 380 °C to 1000 °C. The biochar yield was 49% with an average production rate of 1.8 ± 0.2 kg h−1. The reactor consumed 11 ± 0.1 kg of rice husk as feedstock and 6 ± 1 kg h−1 of wood as fuel. The gaseous products from the pyrolysis were CH4, CO2, H2, CO, and CnHm, which contributed 23.3 ± 2.3 MJ m−3 of energy as fuel for the pyrolysis process. The specific surface area of the biochar was 182 m2 g−1. The achieved operational capacity and thermal efficiency of the reactor show biochar production is a suitable option to convert discarded biomass into a value-added product that can potentially be used in several environmental applications.
item: Article-Full-text
PARROT: Interactive Privacy-Aware Internet of Things Application Design Tool
(2023) Alhirabi, N; Beaumont, S; Llanos, JS; Meedeniya, D; Rana, O; Perera, C
Internet of Things (IoT) applications typically collect and analyse personal data that is categorised as sensitive or special category of personal data. These data are subject to a higher degree of protection under data privacy laws. Regardless of legal requirements to support privacy practices, such as in Privacy by Design (PbD) schemes, these practices are not yet commonly followed by software developers. The difficulty of developing privacy-preserving applications emphasises the importance of exploring the problems developers face to embed privacy techniques, suggesting the need for a supporting tool. An interactive IoT application design tool – PARROT (PrivAcy by design tool foR inteRnet Of Things) – is presented. This tool helps developers to design privacy-aware IoT applications, taking account of privacy compliance during the design process and providing real-time feedback on potential privacy violations. A user study with 18 developers was conducted, comprising a semi-structured interview and a design exercise to understand how developers typically handle privacy within the design process. Collaboration with a privacy lawyer was used to review designs produced by developers to uncover privacy limitations that could be addressed by developing a software tool. Based on the findings, a proof-of-concept prototype of PARROT was implemented and evaluated in two controlled lab studies. The outcome of the study indicates that IoT applications designed with PARROT addressed privacy concerns better and managed to reduce several of the limitations identified. From a privacy compliance perspective, PARROT helps developers to address compliance requirements throughout the design and testing process. This is achieved by incorporating privacy specific design features into the IoT application from the beginning rather than retrospectively
item: Article-Full-text
Exoskeletons for manual handling: A Scoping review
(IEEE, 2023) Perera, S.; Widanage, K. N. D.; Ranaweera, R. K. P. S.; Wijegunawardana, I. D.; Gopura, R. A. R. C.
The prevalence of work-related musculoskeletal disorders is a common issue in many occupations involving manual handling activities. In order to aid manual workers in reducing the burden on the musculoskeletal system, various wearable robotic technologies have been developed over the years. An increase in research work on wearable technologies has been observed, particularly in the last decade. In that context, this article presents a comprehensive review and a bibliometric analysis of the recorded occupational exoskeletons for manual handling since 2010. The review is aimed at identifying the paradigm shifts of research in the recent past and associating the trends pertaining to the applications, mechanisms, and control systems in the development of wearable devices for manual handling. The scope of the review limits itself to active and passive exoskeletons designed to support the upper extremity, lower extremity, and spine for performing load lifting, load carrying, or static holding. The analysis of the results revealed the emerging trends with the aim of providing researchers with areas for improvement and suggestions for different clusters of devices.
item: Article-Full-text
Enhanced sentiment extraction architecture for social media content analysis using capsule networks
(Springer, 2023) Demotte, P; Wijegunarathna, K; Meedeniya, D; Perera, I
Recent research has produced efficient algorithms based on deep learning for text-based analytics. Such architectures could be readily applied to text-based social media content analysis. The deep learning techniques, which require comparatively fewer resources for language modeling, can be effectively used to process social media content data that change regularly. Convolutional Neural networks and recurrent neural networks based approaches have reported prominent performance in this domain, yet their limitations make them sub-optimal. Capsule networks sufficiently warrant their applicability in language modelling tasks as a promising technique beyond their initial usage of image classification. This study proposes an approach based on capsule networks for social media content analysis, especially for Twitter. We empirically show that our approach is optimal even without the use of any linguistic resources. The proposed architectures produced an accuracy of 86.87% for the Twitter Sentiment Gold dataset and an accuracy of 82.04% for the CrowdFlower US Airline dataset, indicating state-of-the-art performance. Hence, the research findings indicate noteworthy accuracy enhancement for text processing within social media content analysis.
item: Article-Full-text
Modeling strength characteristics of basalt fiber reinforced concrete using multiple explainable machine learning with a graphical user interface
(Nature Publishing Group, 2023) Kulasooriya, WKVJB; Ranasinghe, RSS; Perera, US; Thisovithan, P; Ekanayake, IU; Meddage, DPP
This study investigated the importance of applying explainable artificial intelligence (XAI) on different machine learning (ML) models developed to predict the strength characteristics of basalt-fiber reinforced concrete (BFRC). Even though ML is widely adopted in strength prediction in concrete, the black-box nature of predictions hinders the interpretation of results. Among several attempts to overcome this limitation by using explainable AI, researchers have employed only a single explanation method. In this study, we used three tree-based ML models (Decision tree, Gradient Boosting tree, and Light Gradient Boosting Machine) to predict the mechanical strength characteristics (compressive strength, flexural strength, and tensile strength) of basal fiber reinforced concrete (BFRC). For the first time, we employed two explanation methods (Shapley additive explanations (SHAP) and local interpretable model-agnostic explanations (LIME)) to provide explanations for all models. These explainable methods reveal the underlying decision-making criteria of complex machine learning models, improving the end user's trust. The comparison highlights that tree-based models obtained good accuracy in predicting strength characteristics yet, their explanations were different either by the magnitude of feature importance or the order of importance. This disagreement pushes towards complicated decision-making based on ML predictions which further stresses (1) extending XAI-based research in concrete strength predictions, and (2) involving domain experts to evaluate XAI results. The study concludes with the development of a “user-friendly computer application” which enables quick st
item: Article-Full-text
Benefits realization of robotic process automation (RPA) initiatives in supply chains
(IEEE, 2023) Nielsen, IE; Piyatilake, A; Thibbotuwawa, A; De Silva, MM; Bocewicz, G; Banaszak, ZA
Robotic Process Automation (RPA), which automates repetitive, rule-based operations, is becoming a crucial component of today’s enterprises as they compete in more dynamic business contexts. This study intends to provide implications on the Benefits Realization Key Success Factors (BRKSFs) appropriate for RPA projects, given that between 30% and 50% of RPA initiatives fail. The methodology of this study comprises three stages: identify the main contributing BRKSFs, develop a hierarchical relationship model for BRKSFs, and provide real-world examples to show the usability of BRKSFs using two case studies. The results show that having a clear, well-defined, and unchanging process is the most important BRKSF because of its strong influence over other factors. Three factors, namely, aligning the objective of the RPA initiative with the organization’s strategic objectives, choosing the right process for automation, and change management, have lower driving powers but high dependence powers than other factors. The five factors that have both high driving and high dependence powers are: prioritizing the benefits that can be obtained through the RPA initiative, performing a feasibility study, assembling a cross-functional team, having a team leader and receiving support from top management. This study sheds light on the interdependencies between BRKSFs for academics and business professionals, enabling them to determine which variables must be considered most for RPA initiatives.
item: Article-Full-text
Seismic fragility of lightly reinforced concrete school building typologies with different masonry infill configurations
(Elsevier, 2023) Sathurshan, M; Thamboo, J; Mallikarachchi, C; Wijesundara, K
This paper presents the outcome of a research study conducted to establish seismic fragilities of school building typologies in Sri Lanka. The school buildings in Sri Lanka can be characterised as lightly reinforced concrete (RC) buildings, infilled with masonry walls (IMW). However, they are categorised into two typologies based on the structural layouts used (1) Type 1 (T01) and (2) Type 2 (T02). Although, the school buildings can be grouped into two typologies, variabilities in terms IMW configurations and their arrangements are observed among those school buildings. These variabilities in terms of building typologies as well as IMW arrangements were taken to establish seismic fragility curves. The seismic performances of the school buildings were numerically assessed, where in total 640 building cases were analysed by varying typologies, IMW configurations, and stochastic material properties. Since, the RC school buildings are lightly reinforced, a simplified, yet a novel approach was followed to account the shear failure of RC columns under seismic actions. Then, four damage thresholds were established (slight, moderate, severe and collapse) and the corresponding fragility curves are presented in terms of school building typologies considered. Finally, based on the fragility curves, damage probability matrices of the building typologies were established.
item: Article-Full-text
Rapid seismic visual screen method for masonry infilled reinforced concrete framed buildings: Application to typical Sri Lankan school buildings
(Elsevier, 2023) Sathurshan, M; Thamboo, J; Mallikarachchi, C; Wijesundara, K; Dias, P
Seismic rapid visual screening (RVS) methods are used when a large stock of structures is to be evaluated for seismic risk. Although several RVS methods are available, applications of those methods to appraise the seismic risk of reinforced concrete framed (RC) buildings with irregularities in masonry infill walls (MIWs) are limited. School buildings constructed in Sri Lanka are built with certain RC frame typologies; however, they vary in terms of MIW arrangements used. Therefore, a new RVS method is proposed to evaluate the seismic risk of masonry infilled reinforced concrete (RC-MIW) buildings, particularly for the typical RC-MIW school buildings in Sri Lanka. The proposed RVS method incorporates irregularities of MIW arrangements in the typical RC buildings, the attributes of which are not well accounted in the available RVS methods. The vulnerability attributes such as short column and soft storey effects, arise due to the irregularities of MIW arrangements in the buildings, are explicitly incorporated in the proposed RVS method. The FEMA P-154 guidelines were followed to develop basic scores, score modifiers and minimum scores in the proposed RVS method. For that purpose, seismic performances of RC-MIW schools with various MIW irregularities were numerically analysed. The effectiveness of the proposed RVS method is compared with the existing RVS methods to evaluate the seismic risk of typical RC-MIW school buildings in Sri Lanka. It is shown that the proposed RVS method is capable of capturing the seismic risks of such typical RC-MIW Sri Lankan school buildings
item: Article-Full-text
Application of subcontracting as a quality improvement tool for building constructions in Sri Lanka
(Taylor and Francis, 2023) Mallawaarachchi, VT; Senanayake, SMAH; Disaratna, V; Perera, BAKS
Research on the necessity of subcontractors’ incorporation in quality improvement in construction is scarce globally, especially in the Sri Lankan context. Thus, this research aimed to evaluate the applicability of subcontracting as a quality improvement tool for building constructions in Sri Lanka under the traditional procurement path. The qualitative approach with semi-structured interviews using experts for the data collection through three qualitative Delphi rounds was used. The collected data were analysed using manual content analysis. Selection, communication, productivity, time schedules, payments, and safety relating to subcontractors are the key elements to be considered when enhancing subcontractors’ contribution to the quality improvement of building constructions. This study fulfilled the literature gap to evaluate the applicability of subcontracting as a quality improvement tool for building construction, especially in Sri Lanka.
item: Article-Full-text
Investigation of pozzolanic properties of sugarcane bagasse ash for commercial applications
(American Chemical Society, 2023) Prabhath, N; Kumara, BS; Vithanage, V; Samarathunga, AI; Sewwandi, N; Damruwan, HGH; Lewangamage, S; Koswattage, KS
The ideal climatic and environmental conditions for sugarcane cultivation are present all year round in the tropical island of Sri Lanka. Given the annual sugar consumption of the nation, a significant amount of sugarcane bagasse ash (SCBA), a by-product with no intended commercial use but potential environmental and health risks, is produced. Numerous studies have been conducted recently to assess the viability of using SCBA as a pozzolanic material in structural applications. The purpose of this study is to evaluate the microstructure of SCBA samples from three sugar manufacturing facilities in Sri Lanka to identify the pozzolanic capacities. Several quantitative and qualitative characterization techniques have been utilized for the investigations. While maintaining the American Society for Testing and Materials (ASTM) 618 specification as the standard for pozzolanic properties, a comparative investigation of the attributes of samples from each location was conducted. Beyond that, the relationship between the SCBA generation process parameters and their impact on the properties of SCBA have been identified. Finally, the SCBA source of the Pelwatte unit has been identified as the ideal source for the pozzolanic material from the three locations, considering quality and the extent of additional treatments required before use. Other prospective areas of research on SCBA and its potential applications have been recognized.
item: Article-Full-text
Neural machine translation for low-resource languages: A Survey
(Association for Computing Machinery, 2023) Ranathunga, S.; Lee, E.-S. A; Prifti Skenduli, M; Shekhar, R; Alam, M.; Kaur, R.
Neural Machine Translation (NMT) has seen tremendous growth in the last ten years since the early 2000s and has already entered a mature phase. While considered the most widely used solution for Machine Translation, its performance on low-resource language pairs remains sub-optimal compared to the high-resource counterparts due to the unavailability of large parallel corpora. Therefore, the implementation of NMT techniques for low-resource language pairs has been receiving the spotlight recently, thus leading to substantial research on this topic. This article presents a detailed survey of research advancements in low-resource language NMT (LRL-NMT) and quantitative analysis to identify the most popular techniques. We provide guidelines to select the possible NMT technique for a given LRL data setting based on our findings. We also present a holistic view of the LRL-NMT research landscape and provide recommendations to enhance the research efforts further.
item: Article-Full-text
Deep learning based non-intrusive load monitoring for a three-phase system
(IEEE, 2023) Gowrienanthan, B; Kiruthihan, N; Rathnayake, KDIS; Kiruthikan, S; Logeeshan, V; Kumarawadu, S; Wanigasekara, C
Non-Intrusive Load Monitoring (NILM) is a method to determine the power consumption of individual appliances from the overall power consumption measured by a single measurement device, which is usually the main meter. Increase in the adoption of smart meters has facilitated large scale implementation of NILM, which can provide information about individual loads to the utilities and consumers. This will lead to significant energy savings as well as better demand-side management. Researchers have proposed several methods and have successfully implemented NILM for residential sectors that have a single-phase supply. However, NILM has not been successfully implemented for industrial and commercial buildings that have a three-phase supply, due to several challenges. These buildings consume significant amount of power and implementing NILM to these buildings has the potential to yield substantial benefits. In this paper, we propose a novel deep learning-based approach to address some of the key challenges in implementing NILM for buildings that have a three-phase supply. Our approach introduces an ensemble learning technique that does not require training of multiple neural network models, which reduces the computational requirements and makes it economically feasible. The model was tested on a three-phase system that consists of both three- phase loads and single-phase loads. The results show significant improvement in load disaggregation compared to the existing methods and indicate its applicability.