Heca Journal of Applied Sciences

Electrical Resistivity Tomography for Mapping Landslide-Prone Zones in Pining Subdistrict, Gayo Lues, Aceh, Indonesia

Bukhari Ali — 2026-08-25

The interaction of topography, weathering, fracturing, lithological variability, and subsurface moisture controls slope instability in humid tropical terrain. This study maps resistivity heterogeneity and evaluates candidate weak zones on a slope showing evidence of movement in Pining Subdistrict, Gayo Lues Regency, using three intersecting electrical resistivity tomography (ERT) profiles. Data were acquired with a SuperSting R8/IP system using a Wenner-Schlumberger array, 55 electrodes at 6 m spacing, and a 324 m profile length. The apparent-resistivity data were inverted in EarthImager 2D using an L2-norm smooth-model scheme with electrode elevations incorporated into the inversion mesh. Inversion terminated after four iterations for Profiles 1 and 2 and five iterations for Profile 3, with final RMS misfits of 11.62%, 11.41%, and 10.42%, respectively. The inverted resistivity models delineate a near-surface high-resistivity domain (>1000 Ohm.m) overlying intermediate- and low-resistivity domains. Broad spatial correspondence is observed between conductive anomalies S1 and S6, S2 and S7, and S3 and S8 on Profiles 1 and 3, whereas conductive anomaly S5 on Profile 2 is spatially associated with S1 and S6 near the profile intersections. Conductive anomalies S1, S5, and S6 have the strongest cross-profile support. Considering the resistivity distribution, geological setting, field evidence of slope movement, and monthly rainfall context, the principal conductive anomalies are interpreted as intensely weathered, locally fractured, and comparatively moist material. The resistivity-gradient boundary at approximately 10-50 m depth marks a geoelectrical transition and is treated as a candidate weak-zone boundary rather than a confirmed slip surface; hydrogeological and geotechnical validation is required.

RE-PAD: A Modular Renewable-Powered Auxiliary System for Decoupling Radiator Cooling Fans in Isolated Diesel Power Plants

Marzuirman Maizuirman — 2026-09-15

Engine-driven radiator cooling fans in isolated diesel power plants impose a continuous mechanical parasitic load typically reported at 3–7% of rated engine power, a burden compounded by Indonesia's B40 biodiesel mandate, which increases cooling-system heat-rejection demand, yet no study identified in the authors' prior systematic review documents a formal, traceable engineering design process for addressing this problem through auxiliary-load decoupling. This study aims to systematically elicit design requirements, generate and evaluate alternative technical concepts, and specify a traceable modular architecture for parasitic-load decoupling in a B40-fuelled isolated diesel generator. A systematic engineering design methodology, following the Pahl & Beitz/VDI 2221–2223 framework, was applied to the Caterpillar 3512B LY800217 (1024 kW) unit operated by PLN Nusa Daya in Kepulauan Riau, comprising requirement elicitation, functional decomposition, morphological analysis, and Pugh-based concept selection, with requirement coverage examined through a traceability matrix and the robustness of the concept ranking examined through a weighting sensitivity check. Six functional requirements yielded four concept variants; Pugh scoring, supported by the sensitivity analysis, indicated that a fully renewable-decoupled configuration (net score +5) was preferred over a hybrid renewable-with-fallback alternative (net score +3) under equal criterion weighting, principally through complete satisfaction of the parasitic-load-elimination requirement, with the ranking reversing only where retrofit capital cost is weighted dominantly. The selected six-module Renewable-Powered Auxiliary Decoupling (RE-PAD) architecture was shown, through traceability analysis, to conceptually address all elicited requirements without gap, and preliminary literature-based screening indicated technical plausibility relative to comparable variable-frequency-drive and biodiesel-thermal studies; the parasitic-load magnitude (estimated at 41.0 kW, 4.0% of rated output) and the component ratings presented remain preliminary values whose thermal and techno-economic validation is reported in a companion simulation study (in preparation). The architecture supports a conceptually scalable, phased retrofit pathway for the approximately 80-unit PLN Nusa Daya isolated diesel fleet, subject to per-unit sizing confirmation.

Perceived Barriers and Strategies for Enhancing Village Midwives’ Antenatal Care Delivery: A Qualitative Study from Banda Aceh, Indonesia

Tilaili Ibrahim — 2026-09-17

Antenatal care (ANC), a vital part of maternal health services, directly affects the health of both the mother and the newborn. Village midwives are at the forefront of providing ANC to communities in Indonesia. Understanding the strategy to improve village midwives' performance is essential for developing targeted interventions to improve ANC quality and maternal health outcomes. This study aimed to explore midwives’ perceptions of the factors influencing village midwives' ANC service delivery and to identify strategies perceived to improve antenatal care quality in Banda Aceh, Indonesia. The study used a qualitative approach with a maximum variation sampling strategy. Data were collected from October to November 2018 through in-depth interviews with seven informants, including village midwives, midwife coordinators, heads of health centers (Pusat Kesehatan Masyarakat = Puskesmas), and a representative from the Municipality Health Office. The results identified five main themes: some midwives were not domiciled in their assigned villages; midwife coordinators conducted routine evaluations, but feedback was inconsistent; midwives valued continuous learning; maternal and child health records were sometimes incomplete due to limited facilities and irregular visits; and health centers conducted non-random periodic facilitative assessments. Participants proposed several strategies to address these challenges, including internal reviews of the maternal and child health book records, educating pregnant women on the use and importance of the book, standardization of maternal cohort registers, patient satisfaction assessment, and realignment of village midwives’ placement with their official job descriptions, which represent recommendations generated from participants' perceptions rather than interventions tested within this study. These findings offer contextual insight into perceived barriers and improvement strategies for village midwives in delivering ANC, providing a foundation for designing targeted interventions.

Beyond Morphology: COI-Based Molecular Identification of Sea Urchins from Kumu Beach, North Sulawesi

Nur Aisyah — 2026-09-30

Sea urchins (Echinoidea) play important ecological and economic roles in coastal ecosystems, but morphological identification is often unreliable due to overlapping diagnostic traits among closely related taxa. This study used DNA barcoding of the mitochondrial COI gene to identify a sea urchin specimen from Kumu Beach, North Sulawesi, and determine its phylogenetic relationship with GenBank reference sequences. DNA was extracted from gonad tissue, amplified via PCR using the BK-COI-AntF&AntR primers, sequenced, and analyzed through BLAST, MUSCLE alignment, and Maximum Likelihood phylogenetic reconstruction (Kimura 2 Parameter model, 1000 bootstrap replications). PCR yielded a clear ~709 bp fragment, producing a 658 bp COI consensus sequence (A+T content: 56.84%). BLAST analysis showed 100% query cover and identity with multiple Tripneustes gratilla reference sequences, and phylogenetic analysis placed the specimen within the Tripneustes clade, closely nested among T. gratilla sequences. These results confirm the specimen as T. gratilla and support COI based DNA barcoding as a reliable method for sea urchin species identification.

Activity Cliffs and Machine Learning Generalization in QSAR: Why Similar Molecules Remain Difficult to Predict

Teuku Rizky Noviandy — 2026-09-30

Machine learning has substantially expanded the predictive capability of quantitative structure-activity relationship (QSAR) modeling, yet reliable generalization remains difficult in chemically challenging regions. Activity cliffs-pairs of highly similar molecules that exhibit large differences in biological activity-represent a particularly stringent test because they violate the local smoothness assumed by many similarity-based and representation-learning approaches. This narrative review examines why activity cliffs remain difficult for contemporary QSAR models and how they expose limitations that may be concealed by conventional global performance metrics. We discuss the definition, detection, and chemical origins of activity cliffs; their relationship to discontinuities in structure-activity landscapes; and the influence of molecular representation, training-test partitioning, structural redundancy, and data leakage on apparent model performance. Evidence from fingerprint-based machine learning, deep neural networks, graph neural networks, and recent molecular foundation models indicates that increasing model complexity does not consistently resolve activity-cliff prediction. Models with strong overall QSAR performance may still exhibit substantially larger errors or poor sensitivity for cliff-forming compounds. Recent approaches that explicitly incorporate activity-cliff information through pairwise, contrastive, or activity-aware representation learning provide a promising alternative to conventional compound-level regression. We therefore propose an activity-cliff-aware QSAR perspective integrating rigorous bioactivity curation, activity-landscape characterization, leakage-resistant validation, global and cliff-specific evaluation, applicability-domain analysis, uncertainty quantification, and cliff-aware learning objectives. Activity cliffs should ultimately be regarded not merely as problematic outliers, but as information-rich stress tests of molecular machine-learning generalization and local structure-activity understanding.