Authors - Sajidur Rahman, Md. Ataur Rahman, Saymadeen Tabassum, Farhad Alam, Mahady Hasan, Towsif Zahin Khan Abstract - Our work explores the application of the Waterfall model in game development through a case study involving an inexperienced team developing a Space Invaders clone. While Agile methodologies are widely favored in modern game development for their flexibility, the structured, phase-based nature of Waterfall presents certain advantages for teams with limited experience in iterative processes. We investigate how clearly defined stages- requirement analysis, design, implementation, testing, and maintenance-provide direction, minimize scope creep, and can support systematic progress and project success despite the team’s inexperience. Key findings suggest that Waterfall was effective in maintaining project control, with clear documentation and milestone tracking helping reduce ambiguity. However, the model also posed challenges in managing late-stage design changes and balancing creativity within a rigid structure. The paper compares these findings with insights from the literature on Agile and hybrid methodologies, especially in the context of novice teams. The case study concludes that although Waterfall’s predictability can benefit inexperienced developers, its success relies heavily on disciplined planning, well-defined roles, and the ability to adapt within a structured framework.
Authors - Krit Chaiwong, Wirote Jongchanachavawat, Ittipat Roopkom, Boonchart Kativat Abstract - This research and development aimed to create a LINE application to prevent COVID-19 among pregnant women and to evaluate its effectiveness and user satisfaction. The study was conducted in four stages: (1) assessing the situation, needs, related factors, and feasibility of the LINE application; (2) developing the LINE application for COVID-19 prevention; (3) implementing a quasi-experimental study; and (4) refining and evaluating the application’s effectiveness. The research instruments included: (1) in-depth interview questions with content validity index (CVI) values ranging from 0.88 to 0.90; (2) field notes; (3) a questionnaire on COVID-19 prevention among pregnant women, comprising a knowledge assessment (KR-20 = 0.72) and a behavior assessment (Cronbach’s alpha = 0.93); and (4) a satisfaction survey for the LINE application (Cronbach’s alpha = 0.89). Data were analyzed using content analysis, frequency, percentage, mean, standard deviation, and paired t-tests. Results indicated that after using the LINE application, participants demonstrated significantly higher knowledge and improved preventive behaviors compared to pre-intervention levels (p < 0.001). Participants also reported a high level of satisfaction with the application. These findings suggest that the LINE application is an effective and well-received tool for promoting COVID-19 prevention among pregnant women.
Authors - Seema V Kedar, Shabana Pirjade, Avinash Chhabu Shelar, Dheeraj Sharma, Harshita Daftari, Rushikesh Manoj Gholap Abstract - Around 9.76 million peoples in the rural and semi-urban area are depend on the sericulture for their livelihood. Monitoring the health and environment of silk worm is a crucial factor of sericulture industry. Which is done manually by farmers or care taker by monitoring them time to time and keeping track of their environmental conditions which are favourable temperature and humidity. By automating this process lot of time, work and labour cost of farmers can be saved and utilized for other work. For doing so Arduino will be used which will be connected to temperature sensor, and humidity sensor which will be used for taking input signal from the system and define the current environment conditions of the system. The current environmental conditions will be compared by threshold values which will be which will be predefined. If the value of temperature is greater than threshold the fan which is another component will be switched on to cool down the temperature. If the temperature is lower than the threshold the bulb will be switched on to rise the temperature. All the status report of the system will be given send to famers phone by using the GSM module. Throughout the process of monitoring the health of silk worm always had a front foot. Cause one diseased worm can make another worm ill. Hence, to avoid this a disease detection system which will detect diseases silk worm will help farmers. For doing so CNN and Image processing algorithms are used for disease detection on time This model will classify the diseased silkworm in different diseases based on which it is suffering from. CNN is trained with its auto generated features. An average accuracy of 85%, 75% and 59% for classification of healthy and diseased silkworms.
Authors - Parth Modhvadiya, Harshil Kothiya, Nishat Shaikh, Jalpesh Vasa Abstract - The main goal of this paper is to study and examine accident data in India on a national, state, and major city level between 2017 and 2021. It focuses on identifying patterns, key factors, and differences across regions. Road accidents have been a serious issue in India for a long time. Research indicates that injuries resulting from road accidents remain a significant challenge for public health, affecting both the economy and society. These accidents are particularly common in states like Tamil Nadu, Maharashtra, and Uttar Pradesh. The paper uses data from various sources, including government reports, to better under-stand the scale of the problem, regional variations, key risk factors, and possible ways to prevent accidents. One of the main reasons for road accidents in India is careless driving. Furthermore, the rapid increase in vehicle ownership has played a key role in the rising number of accidents. Data from the Ministry of Road Transport and Highways indicates that registered motor vehicles in India grew from 24.8 crore in 2017 to 28.9 crore in 2020, marking an approximate 16.5% rise. This surge in vehicles has caused more traffic congestion, making it harder for drivers to travel safely.
Authors - Hector Rafael Morano Okuno Abstract - Nowadays, Large Language Models (LLMs) are utilized in various fields of knowledge to assist users in performing tasks after being prompted with a question. However, further exploration is needed to determine their scope, as they are constantly being trained to improve and enhance their capabilities. This article investigates an LLM to decide whether or not it can assist users in interpreting results from finite element analyses and whether it can suggest modifications to the analyzed parts to improve the results of said analyses. This work was conducted during the Cyber-Physical Systems course, held in the August-December 2024 semester, with students of mechatronics, robotics, and computational systems engineering. Among the results, it was found that LLMs are capable of recognizing images with beam diagrams featuring different types of support and loads, as well as correctly interpreting finite element analysis results generated by software tool applications, with the potential to provide recommendations for improving these results.
Authors - Hetansh Shah, Hitarth Bhatt, Jay Topiwala, Hitanshu Shah, Pradnya Saval, Shruti Mathur Abstract - This paper compares classical and deep models for forecasting stock prices of India’s top energy stocks (ONGC, NTPC, RELI) under inflation stress measured by the Energy Price Index (EPI). Although classical models like Prophet and Holt-Winters accurately forecast the inflation series, an LSTM network with an Attention mechanism forecasts much better for stock price forecasting, especially for volatile stocks. The paper further demonstrates that pure-energy stocks (ONGC, NTPC) are more inflation-sensitive than diversified RELI. The findings have pragmatic implications for investors and policymakers on risk management in the energy market.
Authors - Stanny Dewanty Rehatta, Meta Bara Berutu, Inkreswari Retno Hardini, Gita Safitri Abstract - The rapid growth of Buy Now, Pay Later (BNPL) services in Indonesia, especially Shopee PayLater, has created new patterns in consumer behavior and technology adoption in digital finance. While these services offer practical benefits and financial flexibility, they also raise concerns related to data privacy, potential misuse of personal information, and system security. This study investigates the factors influencing users’ behavioral intention to use Shopee PayLater by extending the Technology Acceptance Model (TAM). The proposed model includes six key variables: Perceived Ease of Use, Perceived Usefulness, Perceived Risk, Perceived Security, User Trust, and Intention to Use. A quantitative method was applied by distributing a structured online questionnaire to Shopee users in Indonesia. The data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The results show that Perceived Ease of Use positively influences both User Trust and Intention to Use, although its effect on intention is not statistically significant. Perceived Risk negatively affects both Trust and Intention, with a significant impact on Trust. Perceived Security unexpectedly has a negative but marginally significant effect on Intention to Use. User Trust is found to be a strong and significant predictor of Intention to Use, indicating its central role in mediating other variables in the model. These findings suggest that, beyond ease and usefulness, users' trust and risk perceptions significantly affect adoption. Strengthening security assurance and reducing perceived risk are essential strategies to improve trust and drive continued use of BNPL services such as Shopee PayLater in the Indonesian digital finance environment.
Authors - Budsaba Inkiew, Wongpanya S. Nuankaew, Thapanapong Sararat, Pratya Nuankaew Abstract - Tuberculosis (TB) remains a global health threat, especially in low- and middle-income countries like Thailand, where rural areas have limited healthcare resources. This study uses a machine learning–based clustering framework to stratify TB patients and screening data in Ngao District, Lampang Province. It combines data sources—including a registry of 899 TB cases and 15,318 chest X-rays—and addresses challenges like data heterogeneity, missing values, and inconsistent formats through cleaning, imputation, and feature engineering. Using the k-means algorithm, evaluation metrics such as Silhouette Score, Davies–Bouldin Index, and Calinski–Harabasz Index confirmed three optimal clusters. Each cluster shows distinct demographic, clinical, and epidemiological features, highlighting TB risk diversity. The findings demonstrate machine learning's potential to support targeted public health interventions, optimize resources, and improve TB prevention and control. This scalable framework offers insights for other regions facing similar infectious disease challenges, integrating AI into public health systems.
Authors - Aishwini VR, Aravindan B, Santanu Mandal Abstract - The integration of Artificial Intelligence (AI) in education offers unprecedented opportunities for personalization, efficiency, and innovation, yet resistance to adoption persists among Generation Z learners. This study investigates the structural interplay of ten key barriers to AI adoption in higher education, integrating the Technology Acceptance Model, Resistance to Innovation Theory, and Cognitive Load Theory. Using Interpretive Structural modelling (ISM) and MICMAC analysis on survey data from 118 Gen Z respondents, the research identifies lack of trust in AI, data privacy concerns, and algorithmic bias as high-driving factors, influencing dependent barriers such as low perceived usefulness and poor user experience. Digital fatigue and cognitive overload emerge as central linkage variables, mediating relationships between upstream drivers and downstream disengagement. Findings highlight the pivotal role of faculty support and pedagogical integration in shaping adoption attitudes. The study reframes resistance not as reluctance, but as a signal of systemic misalignment between learner expectations, institutional practices, and AI design. Practical implications call for transparent, inclusive, and cognitively considerate AI tools, robust governance frameworks, and AI literacy initiatives to foster equitable and sustainable adoption.
Authors - Prabhuram S P, B Sathish Kumar, Santanu Mandal Abstract - The global fashion industry’s environmental footprint has heightened the urgency for sustainable fashion adoption, yet Generation Z (Gen Z)—despite high sustainability awareness—shows limited engagement. This study investigates the systemic barriers impeding sustainable fashion consumption among urban, digitally active Indian Gen Z consumers. Drawing on the Theory of Planned Behaviour and Resistance to Innovation Theory, the research identifies ten interlinked barriers, including lack of awareness, limited availability, price sensitivity, trend obsession, peer influence, greenwashing confusion, low perceived impact, convenience bias, brand loyalty, and inconsistent messaging. A mixed qualitative–quantitative approach employing Interpretive Structural Modelling (ISM) and MICMAC analysis was applied to 137 valid responses collected in June–July 2025. Findings reveal that lack of awareness and limited availability are foundational drivers shaping intermediate constraints—such as price sensitivity and low perceived impact—which ultimately influence entrenched behaviours like brand loyalty and trend fixation. The ISM hierarchy underscores the multi-level nature of resistance, while MICMAC classification highlights critical driving factors for targeted interventions. The study contributes to sustainable consumption theory by mapping structural interdependencies and offers actionable insights for policymakers, educators, and brands seeking to align Gen Z fashion choices with Sustainable Development Goals 12 and 13.
