Harumi Veny

Sustainable water resource management is an increasingly urgent global challenge, where conventional methods are often inadequate. Artificial Intelligence (AI) has emerged as a transformative technology offering advanced solutions. This paper presents a systematic review of the progress, applications, and challenges of AI in water management during the period 2010–2025. Through a thematic analysis of relevant literature, we identify three distinct evolutionary stages: an early stage dominated by traditional machine learning (2010–2015), a deep learning revolution (2016–2020), and an era of Advanced AI Integration and Innovation featuring hybrid models and physics-aware machine learning (2021–2025). Key findings show that AI excels in various applications, particularly high-accuracy water quality prediction, real-time monitoring systems, process optimization, and predictive analytics for disaster mitigation. Despite its significant strengths in accuracy and data processing, major challenges remain, including data availability limitations, lack of model interpretability (“black box”), and generalization difficulties. This review concludes that future research directions, such as Explainable AI (XAI) and domain knowledge integration, are crucial to overcoming these barriers and realizing the full potential of AI in creating intelligent, efficient, and resilient water management systems.

Efficient light rail transit (LRT) systems are crucial for sustainable urban mobility; however, unforeseen departure delays continue to be a major hurdle, undermining operational reliability and passenger satisfaction. This study establishes a data-driven framework for forecasting departure delays by combining static GTFS schedules with real-time GTFS operational data from the Canberra LRT system. The dataset included 15,538 records with 42 attributes, spanning from 28 August 2020 to 13 August 2022. A stringent preprocessing pipeline was implemented, encompassing temporal feature engineering and feature selection based on mutual information. The Random Forest regressor with feature engineering and selection (RFR-FEFS) attained the highest predictive performance on the test set (R 2 = 0.94, MAE = 2.93, MSE = 34.32). The high accuracy indicates the model's efficacy, yet it necessitates careful evaluation of potential overfitting and its generalizability beyond the examined system. Ablation experiments were performed to assess the impact of various feature groups by omitting temporal, spatial, or operational attributes. The findings indicate that the exclusion of temporal features decreased R 2 to 0.90, the exclusion of spatial features reduced it to 0.93, and the exclusion of operational features resulted in the most significant decline to 0.23. These findings affirm that all three feature categories contribute distinctly and synergistically to model performance. This research illustrates the capability of integrating diverse GTFS data with sophisticated machine learning techniques to attain precise LRT delay forecasts. Nevertheless, the framework was validated solely on one system and time frame; future research should investigate its transferability to other cities and integrate supplementary contextual data, including meteorological conditions and incident reports, to improve robustness and practical applicability.

This study explores the effectiveness of micro-learning modules in enhancing student engagement, understanding, and motivation in chemical engineering, specifically in teaching complex concepts related to separation processes. This study successfully collected 100 respondent data through surveys focusing on seven key areas namely pace of learning, engagement, retention of information, clarity of material, preference of learning, motivation, and concept understanding. The findings revealed a strong positive response towards micro-learning. The overall average ratings were high, with engagement rated at 4.13 (SD = 0.73), clarity of material at 4.12 (SD = 0.74), and pace of learning at 4.11 (SD = 0.75), showing a slight lead in engagement. The study concludes that micro-learning is highly effective in enhancing student engagement, understanding, and motivation in chemical engineering education. These findings support the continued integration and development of micro-learning strategies to improve student outcomes, particularly in fields requiring mastery of complex concepts. Further research is suggested to explore enhancements in retention and motivation to maximize the benefits of micro-learning.

Cosmetics have progressed into an essential part of daily routines due to their widespread use, commonly through topical application. However, many cosmetic products contain heavy metals which can accumulate and pose health risks over time. Concerns on product safety including ingredient monitoring have been raised as regular cosmetic use exposes consumers to potentially harmful chemicals. Understanding consumer perceptions is crucial for informed decision-making, yet research on user perception of cosmetics and heavy metal contents in the Malaysian market remains limited. This study focused on assessing heavy metal contents in face cream products from various brands as heavy metals pose a significant concern. The concentration of heavy metals in cosmetic products was analysed by using inductively coupled plasma-optical emission spectrometry (ICP-OES) and atomic absorption spectroscopy (AAS). The results indicate that 4440 ppm of mercury is present in Brand B, which is particularly alarming. These findings underscore the prevalence of unauthorised local products with higher-than-permissible levels of heavy metals, highlighting the need for stricter monitoring and regulation.

The vertical deviation is one of the important stages in commercial aircraft flights because indications of malfunction can be detected faster by the pilot rather than in other flight phases. The purpose of this study is to investigate whether the PK-LQP airplane experienced unusual altitude movements during the vertical deviation phase since the airplane taking off from Seattle (Boeing Manufacture's home base) until the airplane crash. The airplane, which is of the type B737 MAX8, had operated for 83 days, during which it completed 438 flights using a total of 107 flight codes, and it travelled to 36 airports. According to the findings of an investigation of the data, we found that only 69 (39%) were included in tier 1, which had an ADS-B data update interval message below 10 seconds, complying with ICAO and FAA-AVN standards. With this class 1 and tier 1 dataset, we conducted an EDA to find the data insight, which revealed that there was a disturbance in the speed and altitude indications on the airplane instrumentation, causing a misperception for the pilot and causing the airplane to drastically drop in altitude (more than 100 feet).

Biodiesel, a third-generation bio-fuels, offering several advantages over regular diesel fuel. Waste cooking oil (WCO) emerges as an ideal feedstock due to its availability and easy accessibility. In this work, biodiesel is utilized from two different types of immobilized lipases: Rhizomucor miehei lipase (RMIM) and Candida antarctica lipase B (CALB). The impact of the molar ratio of oil to methyl acetate (1:3-1:12) was evaluated for both lipases, and the resultant biodiesel was tested in diesel engine. The enzymatic transesterification was carried out in ultrasonic assistance and the results showed that the greatest yield of 81.20% at 45℃, using CALB as a biocatalyst, 1.8% (w/v) lipase and oil to methyl acetate molar ratio of 1:12 within 3 hours. Triacetin, by-product was determined their concentration for each molar ratio and analyzed using FTIR range of 500cm-1 to 4000cm-1, revealing a significant absorption peak at 1238.90cm-1. Biodiesel was blended with commercial diesel fuel in varying quantities of 7, 10, and 20% by volume (B20). The results were compared to Industrial Diesel Fuel 7% (B7) and Commercial Diesel Fuel 10% (B10). NOx and CO2 emission drops as the percentage of diesel/biodiesel blends increases, supporting WCO as a cost-effective biodiesel feedstock with low petrol pollution.

In the present study, biodiesel was produced via the enzymatic interesterification of Crude Palm Oil (CPO) and methyl acetate within ultrasonic condition. In contrast to alcohol, methyl acetate as an acyl acceptor does not inhibit lipase activity and can create triacetin as a useful byproduct. In this work, Immobilized lipase from Candida Antartica A (CaLA) was utilized as biocatalyst and the effect of using non-pretreated CPO and pre-treated CPO as feedstock were explored. The pre-treatment of CPO involves degumming with acid, washing with water, and bleaching. The enzymatic interesterification was conducted in three-neck flasks using an ultrasonic water bath at 45o C. Few parameter effects on biodiesel production were also investigated, including the effect of molar ratio of CPO to methyl acetate, the effect amount of lipase, and the reusability of immobilized lipase (CaLA) in the interesterification reaction. The highest average Biodiesel yield of 80.6% was obtained from pretreated CPO at a molar ratio of 1:9 with 100 mg (1% w/v) of Immobilized CaLA, after three hours of reaction. Further research on the reusability of immobilized CaLA revealed that the yield of biodiesel reduced significantly after the second run. The results of the present study also demonstrated that Immobilized CaLA performed well at low concentrations but had low stability, with productivity decreasing to 92% upon reuse after the initial run. In order to make Immobilized lipase economically viable, further research must be conducted to overcome its low stability in the reaction.

Amid the global COVID-19 pandemic, movement restrictions were implemented by various countries, Indonesia being among them. In this study, Jakarta, the capital city of Indonesia, serves as a focal point to elucidate the influence of such restrictions on air quality. Data, spanning from January to April 2020, were obtained from reputable sources including the Indonesian Meteorological Agency, Air Quality Index (AQI) platforms, IQAir, and Jakarta’s open data repository. Upon meticulous analysis, it was observed that diminished human activities in Jakarta resulted in a notable enhancement of air quality, with a variance of 19 days of superior air quality compared to the previous year. For a comprehensive understanding, comparative analyses with neighbouring cities were undertaken. In Singapore, reductions ranging from 8% to 43% were discerned in PM10, PM2.5, CO₂, and SO₂ levels. Bangkok reported a decline of 22% in PM2.5 concentrations, whereas Kuala Lumpur exhibited reductions in PM2.5 levels, varying between 3% and 35%, with further reductions of 3% to 63% in NO₂ concentrations. Concurrently, SO₂ levels saw a decrease of 2% to 48%, and CO concentrations diminished by 1% to 27%. The findings underscore the dual benefits of mobility restrictions, addressing both the containment of COVID-19’s proliferation and substantial improvements in environmental air quality.