4th Annual Research Conference of SLIIT CITY UNI

Permanent URI for this collectionhttps://rda.sliit.lk/handle/123456789/4160

Browse

Now showing 1 - 20 of 33

Open Access
Adaptive Multi-model Machine-Learning and AI Systems for Strengthening the Emotional Well-being of Children with Trisomy 21
(SLIIT City UNI, 2025-07-08) Balasuriya, M.I.D.C.; Ellepola, N.
This research study demonstrates a web application designed to strengthen children's cognitive skills and emotional well-being with Trisomy 21, utilizing interactive and tailored tools. Trisomy 21 is a chromosomal anomaly caused by an extra copy of the 21st chromosome, which affects a child's cognitive development. Despite the technological evolution, a significant gap persists in accessibility and multimodal approaches that meet the unique needs of children with Trisomy 21. The main objective of the research study was to develop a multi-model web-based application, “Mockley kids,” customized for children with Trisomy 21 that helps to enhance their cognitive skills and emotional wellbeing. The developed system integrated an artificial intelligence (AI) powered voice assistance to enhance communication, and learning, an emotion-based music recommender to enhance emotional well-being and provide a calm and uplifting environment, a text-based bot to enhance literacy skills and communication, and interactive games “Who am I?”, “Tic-tac-toe”, “Simon-says” to increase attention span, decision making, which are tailored to enhance their cognitive skills and emotional well-being. The development and implementation of this project follows a structured process aligned with Agile project methodology. To evaluate the “Mockley kids” system’s impact on children with Trisomy 21, 16 children who were diagnosed with Trisomy 21 engaged with the system for 20 minutes for a week, under the supervision of 8 professionals, including 3 speech therapists, 3 occupational therapists, and 2 educators. Overall results show that the children were excited to integrate with the system and enjoyed the system. Both the professionals and the parents stated that they had evident noticeable improvements in cognitive abilities, including enhanced communication, memory recall, enhanced attention span, and improvements in emotional well-being.
Open Access
AI-Based Smart Traffic Management System for Emergency Vehicles
(SLIIT City UNI, 2025-07-08) Amarasinghe, D P S V; Benorith, L
Modern cities' main traffic congestion problem delays emergency vehicles like ambulances and firetrucks and police cars where every second counts. Fixed signal traditional traffic systems lack real-time adaptability, hence delays and risks are raised. This paper suggests an AI-driven smart traffic management system to priorities emergency vehicles and enhance general traffic flow by means of Raspberry Pi, YOLOv8, and OpenCV. Strategically positioned cameras provide video to a Raspberry Pi, which detects emergency vehicles by using OpenCV and YOLOv8. Dynamic control of traffic lights on detection helps to clear the path, so reducing response times and improving safety. The technology also maximizes road use and helps to ease traffic. For cities with limited infrastructure, using reasonably priced, open-source tools are scalable and ideal.
Open Access
AI-Driven To-Do List: Optimizing Task Categorization and Prioritization Using Ensemble Models
(SLIIT City UNI, 2025-07-08) Vishaliney, P.; Pemasiri, C.S; Kanthakumar, M; Yatigammana, N
This paper introduces the AI-driven smart todo list that can cluster and prioritize activities by using the machine-learning methods. Traditionally, to-do list services are immovable and have an element of compromising users to input the information themselves; this sort of bare tool can easily lead to unproductiveness in accomplishment of duties. To address this situation, we supplement ensemble modeling, namely Logistic Regression, XGBoost, and Multilayer Perceptron, to delegate the tasks to the desired categories and define priorities by their urgency. Measured based on standard measures, the ensemble will achieve 47.7 percent accuracy when doing classification and 72.8 percent when predicting priority, and High Priority tasks will gain in this evaluation. Using BERT-based embeddings in combination with TF-IDF-based vectorization, the system should improve its effectiveness because it understands the semantics of described tasks. Together these blocks form a superb ensemble architecture that can beat stand-alone model when it comes to classification and forecasting. More importantly, the system still leaves itself potential to adjust to user behavior and therefore it can improve task management, and it is a feasible platform in real time organization of tasks.
Open Access
An AI-Powered Web Application for Waterfall Recognition and Eco-Tourism Enhancement in Sri Lanka: Falls Explorer
(SLIIT City UNI, 2025-07-08) Ranasinghe, S; Jayaweera, Y
This research presents the development of Falls Explorer Sri Lanka, a mobile-responsive web application that uses artificial intelligence for automatic waterfall recognition. The core innovation lies in applying a custom-developed convolutional neural network (CNN) to classify waterfall images based on their visual features. A custom image dataset was created by collecting and organizing photos of popular waterfalls in Sri Lanka, and the model was trained using TensorFlow. The custom CNN model achieved 92% validation accuracy after 25 epochs of training, with inference times under 1 second per prediction. The system successfully classified waterfall images across 20 different waterfall classes with precision scores ranging from 88% to 95%. Users upload a photo of a waterfall through the interface, and the system returns the predicted waterfall name along with travel details from a local JSON database. In addition to the recognition feature, the platform offers comprehensive functionalities such as displaying detailed waterfall information (name, location, description), listing nearby hotels, showing current weather forecasts for safe travel planning, hosting a community forum for users to share experiences and images, providing a carbon footprint calculator to estimate travel impact, and an interactive location search map to explore specific sites manually. This solution bridges the gap between technology and ecotourism, supporting conservation-friendly tourism by enabling travellers to appreciate natural attractions without invasive markers or infrastructure.
Open Access
AirPoint Lab: AI-powered Online Car Painting Customization and Estimation Platform
(SLIIT City UNI, 2025-07-08) Nanayakkara, H.K; Nallaperuma, P
The AirPoint Lab project introduces an innovative AI-based online platform designed to modernize car painting services in Sri Lanka, addressing key challenges such as inaccurate cost estimations, unreliable workshop selection, and limited customization options. The platform features an intuitive web interface that enables users to design and visualize custom car paint jobs with the assistance of AI-driven color matching and instant cost estimates. By integrating workshop ratings, local service recommendations, and transparent pricing, the platform enhances accessibility and trust for users. Developed using Agile Methodology, the system was initially built on the MERN stack but later migrated to WordPress improved scalability, while incorporating machine learning for personalized AI recommendations. Rigorous testing confirmed the platform’s functionality, usability, and accuracy, demonstrating its potential to streamline the car painting process for both customers and service providers. Beyond its practical applications, this web application bridges academic research with realworld implementation, offering a scalable solution adaptable to developing nations.
Open Access
Bat.CG: Development of a Customizable Cricket Character Generator Web Application for Enhanced Broadcasting Experience
(SLIIT City UNI, 2025-07-08) Kiriyalagammana, P. P; Niranga, G.D.H
Cricket broadcasting has evolved significantly with technological advancements, yet traditional systems remain fragmented and technically complex for broadcasters. This research presents Bat.CG (Cricket Character Generator), an innovative web-based application that integrates customizable broadcasting graphics with realtime scoring, neural network-driven predictions, and enhanced audience interaction. The system addresses critical gaps in existing broadcasting infrastructure by providing a unified platform that eliminates the dependency on separate Character Generator (CG) and ball-by-ball scoring systems. Through comprehensive market research involving 81 industry professionals, this study identified key requirements including customizable graphics (99% demand), emergency score updating capabilities (87.7% essential), and integrated CG-scoring systems (49.4% extremely valuable). The proposed solution utilizes the MERN stack (MongoDB, Express.js, React, Node.js) architecture with hybrid neural network and regression models for match predictions. Key innovations include drag-and-drop graphic customization without programming knowledge, real-time data synchronization with sub-second latency, role-based access control, and interactive viewer engagement features. The system's modular design ensures scalability, security, and accessibility while maintaining professional broadcast quality. This research contributes to democratizing cricket broadcasting technology and establishing a foundation for future sports media innovations in developing regions.
Open Access
Beyond the Wrist: Holographic Pathway for Universal Depression Management
(SLIIT City UNI, 2025-07-08) Weerasuriya, B.M.; Egodage, M.D; Ranasinghe, R.K.N.N; Vithurshika, J.; Vihansa, N.K.V.; Niranga, G.D.H
This concept paper introduces a novel smartwatchbased system that leverages artificial holographic technology to address the growing need for accessible mental health support, particularly for individuals experiencing depression. Recognizing the communication barriers and lack of resources for the deaf community, the proposed system is designed to be inclusive for both deaf and non-deaf users. This system blends artificial intelligence, holographic technology, mood tracking, and an inventive smartwatch that can detect individual emotions. A smartphone application will be used to oversee and control each of these components. By integrating wearable technology with emotional wellbeing support, the proposed model will provide continuous, accessible, and user-friendly assistance. If implemented, this tool could enhance user engagement and emotional awareness in therapeutic contexts. To validate feasibility and effectiveness, further research and development are needed.
Open Access
Bliss2Glamour: An Artificial Intelligence Integrated Educational Platform for Skincare and Beautician Training
(SLIIT CITY UNI, 2025-07-08) Ranthatige, N; Seneviratne, O
This research paper represents Bliss2Glamour, an artificial intelligence (AI) based educational platform developed to assist all the NVQ Level 4 blooming beauticians, qualified lecturers, and beauty enthusiasts. Bliss2Glamour has an integrating Learning Management System (LMS), well trained AI chatbot, Selfaffirmations to keep the users motivated, calming music for the salon purposes, provide 24/7 skincare consultation from a highly qualified cosmetologist via WhatsApp website for beautician training, standard online quizzes for the trainee beauticians to get prepared for the exam aligned with the TVEC syllabus. This research paper highlights the motivation, methodology (Agile), implementation (FastAPI, React, React JS Query, fine-tuned QWEN 2.5 0.5B-Instruct AI model), and evaluation (via Weights & Biases). All objectives were met. The AI chatbot achieved 80% accuracy rate based on evaluation using Weights. These results confirm that Bliss2Glamour successfully combines educational content, AI technology, and holistic care into one user-friendly system.
Open Access
CacheBook: A Smart Personal Finance Tracker with OCR-Based Expense Logging and Visual Reporting
(SLIIT City UNI, 2025-07-08) Perera, P.T.; Vithana, V.N
“CacheBook” is a mobile application designed to simplify personal finance management by integrating real time expense tracking with automated receipt scanning. Built using React Native and Firebase, the application enables users to log daily expenses, set financial goals, and visualize spending through reports. A key innovation of “CacheBook” is the use of Optical Character Recognition (OCR) technology, which allows users to scan physical receipts by reducing manual entry and enhancing data accuracy. The app categorizes expenses, provides visual summaries via charts, and alerts users when spending goals are exceeded.
Open Access
Computer Vision Controlled Humanoid Robotic Arm
(SLIIT City UNI, 2025-07-08) Firdouse, M S; Benorith, L
This paper presents the design and implementation of a low-cost, vision-based gesture-controlled humanoid robotic arm that mimics human hand and wrist movements in real time. The system uses a USB webcam and MediaPipe for hand landmark detection, OpenCV for image processing, and a Raspberry Pi 4 to compute landmark vectors and control servo motors via a PCA9685 driver. Calibration modes were introduced for each joint to ensure accurate servo mapping. The solution supports full gesture-based manipulation of a five-fingered robotic hand, including wrist orientation, with minimal latency and no physical contact. The system provides a more intuitive and natural method for robotic arm control compared to traditional input devices and has potential applications in prosthetics, automation, and human-robot interaction.
Open Access
Decentralized Database Management: A Comprehensive Review of Blockchain- Based Data Systems
(SLIIT City UNI, 2025-07-08) Sukirthan, T; Arunpirakash, S; Karuneswaran, G; Tharmmendra, T; Vithiyasahar, V
The emergence of blockchain technology has revolutionized decentralized data management by offering robust alternatives to traditional centralized database systems. This paper provides a systematic and comprehensive review of blockchain-based distributed databases, highlighting key architectural transformations, core enabling technologies such as Merkle Trees, PBFT, and Zero-Knowledge Proofs, and comparing them with conventional distributed databases. Real-world implementations including Hyperledger Fabric, BigchainDB, and OrbitDB are analyzed to assess their scalability, interoperability, and security capabilities. The paper also explores intrinsic security mechanisms, performance bottlenecks, and regulatory challenges that affect adoption. Finally, it identifies open research questions and future directions necessary for building scalable, privacy-aware, and interoperable decentralized database ecosystems suitable for enterprise and multi-stakeholder environments.
Open Access
Design & Implementation of Smart Waste Collection System with Optimized Route Planning
(SLIIT City UNI, 2025-07-08) Josheph, A L; Rajaguru, N M
With the growth of the global population, the challenge of waste management has become more complex. Effective waste management is no longer just a necessity but a critical concern for public health, urban sanitation, and ecological sustainability. Identifying the filled public waste bins in cities and collecting them effectively is one of the essential initial tasks of waste management, to address it more effectively and cost-efficiently the "Smart waste collection system with optimized route planning" is designed. The prototype system utilizes microcontrollers, a Photo- Voltaic power supply unit, ultrasonic sensors, and a wireless module to detect and communicate bin status in a multi-hop wireless network cluster. Each bin is equipped with an Ultrasonic sensor to measure the garbage level and the data is transmitted through neighboring bins to a central control unit using the ISM-based wireless communication multi-hop network, enabling remote monitoring. When bins are detected as filled, their locations are processed to determine the shortest and most efficient route for the collection of garbage using the Google Maps Directions API. The system was tested successfully using a prototype with three waste bins equipped with sensing units, and setting the location on the maps through a cloud-based web visualization built using basic HTML, Python, and Javascript, the system performed accurate fill-level detection and real-time route planning for garbage collection. This solution offers a cost-effective and scalable approach to smart urban waste management by reducing manual monitoring and optimizing collection routes with traffic awareness.
Open Access
Determinants of Adoption of Artificial Intelligence for Business Sustainability: A Study of Small Businesses in Jaffna
(SLIIT City UNI, 2025-07-08) Inthu, M; Jeyaramanan, S; Nimalathasan, B
This study aims to examine the Determinants of the adoption of Artificial Intelligence for Business Sustainability in small businesses in Jaffna. In that context, a deductive approach is employed by the researcher, and data from 72 small businesses in Jaffna. The purposive sampling was used by the researchers. In that context, the data were collected for a business that has potential for digital innovation. As the researcher employed a purposive sampling method, it is ensured that participants who could meaningfully respond to AIrelated questions were selected, given the region’s varied levels of technological awareness. The closed-ended, structured Likert Scale questionnaire was used by the researcher to collect the data. Awareness of AI adoption, Digital Skills, technological infrastructure, and perceived economic benefits are considered independent variables of the study, while business sustainability is considered a dependent variable of the study. The researcher used the SPSS package to analyze the data and the Kolmogorov– Smirnov test, reliability test, correlation analysis, Variance Inflation Test run by the researchers. Among the predictors, the perceived economic benefits (B = 0.395, p = 0.001) have a statistically positive relationship with business sustainability. Followed by Awareness of AI Adoption, which has a statistically positive relationship with business sustainability (B = 0.312, p = 0.004). Furthermore, Digital skills (B = 0.271, p = 0.015) and technological infrastructure (B = 0.219, p = 0.037) also have statistically significant positive effects, highlighting the importance of digital capacity. Additionally, perceived economic benefits have a statistically positive relationship with sustainable business practices.
Open Access
Developing AI-Powered Android Application about Self-financial Management for Individuals “FinGuard”
(SLIIT City UNI, 2025-07-08) Samarakoon, S.M.A; Nallaperuma, N.A.P
In this paper the author presented the development of “FinGuard” an artificial intelligence powered android application intended to aid individuals in effectively managing their finances. The application addresses common money issues like executive daily expenditures, lack of income, and financial ignorance. They lead to individuals taking loans, pawning items or borrowing money actions that compromise long-term financial stability. FinGuard offers income and expense tracking services, automated report generation, monthly predictive analysis, customizable reminder feature, advice for financial management feature and a chatbot for get answers for user problems inside the application. The user credentials are shielded from unauthorized use through secure login functionality. FinGuard applies a full and smart process to improve personal financial wellness and promote good financial management habits.
Open Access
Emily Brontë’s ‘Sense of Place’ as Portrayed in Her Literary Works
(SLIIT City UNI, 2025-07-08) Ponnamperuma, P
‘Sense of Place’ is a theory that defines the emotional attachment individuals develop with specific locations, encompassing both positive and negative feelings. It can be directly applied to the analysis of the Victorian writer Emily Brontë’s portrayals of her house in Yorkshire moors where she grew up. It is reported that she loved her house, and it provided her with a constant back drop for her imaginative thinking and creative writing. It is assumed that the landscape seen through its windows and the sounds heard while being inside it frequently inspired most of the locational portrayals in her poetry and fiction. In consideration of Emily Brontë’s romantic attachment to her home, the present study intends to explore her own sense of place as portrayed in her works. Accordingly, it pursues the research questions, “How is Emily Brontë’s sense of place portrayed in her works, ‘Wuthering Heights’, ‘Remembrance’ and ‘Fall, leaves, fall?’ The methodology involves a thematic analysis of her biography by Edward Chitham, ‘A Life of Emily Brontë’ under the three themes, ‘Emily Brontë’s sensitivity towards nature and the environment,’ ‘Emily Brontë’s emotional intensity,’ and ‘her imaginary encounters with recurring patterns of the orphans and abandoned characters.’ The findings of the present study foreground that Emily Brontë’s was heavily influenced by her romantic perception of the beauty of nature and her identification of the therapeutic power of nature as portrayed through her characters. According to her, nature tends to improve the quality of relationships among humans while showing the ways in which their deterioration starts due to the imbalance of their emotions including the sense of being abandoned.
Open Access
Enhancing Pronunciation Proficiency via Listening Practice for Specific information with the help of ELSA Speak: Evidence from Foundation-Level Students at SLIIT City Uni, Sri Lanka
(SLIIT City UNI, 2025-07-08) Thennakoon, P; Marasinghe, A; Sadithma, M; Gunaratne, N; Mendis, R; Nilaweera, I
This study investigates the impact of English Listening skills for specific information (ELSSI) on English pronunciation (EP) among 40 Foundation semester 2 students at SLIIT City Uni. A mixed approach of questionnaires and experimental tests as research instruments, along with the AI tool, ELSA Speak, was applied to this group. Participants received structured, listening-based practice over the past few days by employing the ELSA Speak AI application. The Data were analyzed by using the Statistical Package for the Social Sciences (SPSS) and Thematic analysis. Therefore, the results show a statistically significant improvement, based on the means of the inclusive posttests (83.49) being higher than the pretests (77.20), highlighting the effectiveness of targeted listening activities in enhancing pronunciation. The increase in the mean score was discovered after the listening treatments by the researchers. An increase in student scores was observed following the treatment. Positive feedback of the sample demonstrated the above-mentioned points. These results lead to the suggestion that ELSSI can help with the development of oral proficiency, while mobile-assisted language learning tools (ELSA Speak) play a crucial role when focusing on the improvement of oral proficiency, further offering valuable insights for educators and language enthusiasts around the world.
Open Access
Explainable AI Powered Mental Health State Capturing Application to Support Students’ Mental Wellness and Academic Stress Mitigation
(SLIIT City UNI, 2025-07-08) Welarathna, J.H; Nallaperunma, P.
Mental health is a state of well-being that enables individuals to manage stress, work effectively, and contribute to society. However, reports show that serious mental health problems among students worldwide are increasing rapidly. A critical problem is that students often fail to recognize mental health issues or the sources of their academic stress, leading to silent suffering that escalates over time. A significant research gap exists as current assessments methods lack the ability to identify root causes of academic stress and provide explainable decisions for clinical use. This significant rise in many students’ mental health issues have indeed opened important discussions about its underlying causes, consequences, and the need for a comprehensive support system. Voices are an important part for identifying emotional expressions, as speech is the most vital channel of communication, enriched with emotions. The system analyzes emotional patterns in students' voices using Natural Language Processing (NLP) techniques to identify eight emotions and reveal the root causes of their mental health challenges and academic or non-academic stress. Additionally, Explainable AI (XAI) techniques are employed to provide a comprehensive analysis of these patterns, enhancing understanding and supporting managerial decision-making. The system achieves 93.46% accuracy using Random Forest algorithm with reliable confidence levels for clinical applications. It operates effectively in uncontrolled environments with language-independent features, ensuring adaptability across diverse student populations. While students typically seek support from counselors and healthcare professionals who base their decisions on clinical experience, this system offers an additional diagnostic tool to complement and validate professional evaluations. This research aims to better understand student mental health issues and contribute to improved students’ wellness and academic success.
Open Access
Exploring Sustainability-Driven Fintech Usage Intentions Among Gen Z in Sri Lanka
(SLIIT City UNI, 2025-07-08) Vallaven, L
With growing global environmental awareness, Generation Z (Gen Z) has emerged as a key demographic driving demand for sustainable, technologyenabled solutions. This concept paper, developed from an undergraduate research proposal, explores sustainabilitydriven intentions to use financial technology (FinTech) services among Gen Z in Sri Lanka. Drawing on the Theory of Planned Behavior (TPB), the study aims to examines and offer new insights on how attitudes, subjective norms, perceived behavioral control (PBC), and environmental concerns influence sustainability-driven Fintech usage intentions among Gen Z’s in Sri Lanka. A mono-method quantitative design is proposed, with data to be collected from Gen Z individuals in Sri Lanka using purposive and snowball sampling. The data will be collected via an online questionnaire and analyzed using descriptive and inferential statistics.
Open Access
Fitness Warrior: Fitness and Nutrition Tracker with Personalized Goal Generation
(SLIIT City UNI, 2025-07-08) De Mel, Y.D; Nallapperuma, P.M
Fitness Warrior is a comprehensive mobile fitness tracking application developed using React Native and Firebase that addresses critical limitations in existing solutions through the innovative integration of machine learning, gamification, and social features. Traditional fitness applications suffer from inaccurate step detection (with error rates exceeding 20% error rates), inefficient nutrition tracking interfaces, poor user retention (with 73% abandonment within three months), and a lack of adaptive personalization. This project uniquely implements ondevice machine learning via TensorFlow.js for privacypreserving step detection, combines TF-IDF vectorization with cosine similarity for efficient food searching, and incorporates principles of Self-Determination Theory through a cohesive social motivation framework. Development followed the Agile Scrum methodology, implementing a CNN-based model processing sensor data at 50Hz sampling rate, creating a database of 2,395 food items with optimized search algorithms, and designing gamified social features. The application achieves 95.2% real-world step counting accuracy compared to manual counting, significantly outperforming conventional threshold-based approaches (48.3% accuracy), while the calorie tracker delivers 92.7% relevant results in top-5 suggestions with 126ms search latency. Evaluation with 21 users demonstrated exceptional impact: 95.3% reported increased daily steps, 90.4% experienced greater calorie intake awareness, and 71.4% found social features strongly motivating. The application received outstanding approval with 90.5% of testers rating overall satisfaction at 8 or higher on a 10-point scale. This research successfully demonstrates how integrated, machine learning-enhanced fitness applications can meaningfully impact user health behaviours while overcoming significant limitations in existing solutions.
Open Access
FocusBoost – A Study Aid with Adaptive Learning Techniques
(SLIIT City UNI, 2025-07-08) Prabaharan, N; Dampalessa, D.R.C.G.K.
FocusBoost is an AI-powered adaptive learning platform designed to support children with Attention Deficit Hyperactivity Disorder (ADHD) through personalized learning experiences. By integrating video-based learning with voice input analysis, the system uses speech processing techniques to assess a child's engagement and comprehension in real-time. Based on real-time analysis, the platform dynamically adjusts content difficulty and pace to the needs of the individual learner. In practical testing, the system demonstrated high accuracy in classifying learner engagement and comprehension, with more ADHD learners reporting improved focus and content retention. Additionally, parents have noticed positive changes in their child’s study habits and attention span through its use. The site has a performance tracking accuracy page for children, which shows their level of comprehension. This research highlights the effectiveness of AI-enhanced learning for students with brain and neurological issues and its potential to improve inclusive, sustainable education practices. The system is designed with scalability in mind, allowing for multilingual support, culturally adaptive content, and future integration with medical professionals, expanding its impact across a variety of educational and therapeutic settings.