Journal La Multiapp

Efficient Detection of Anomalous User Behavior in Cloud Environments Using Simulated Mouse Dynamics and a Hybrid Ensemble Model

2026-09-17T06:42:20+00:00

This study proposes a lightweight and interpretable framework for identifying suspicious behavior within a cloud-based environment through the use of simulated mouse dynamics and a voting ensemble classifier. Contrary to existing approaches that heavily focus on the use of deep learning techniques, this study utilizes a combination of three classical machine learning classifiers: XGBoost, Random Forest, and Support Vector Machine (SVM). The proposed framework utilizes a voting classifier that achieved a test accuracy of 51.1%, effectively identifying 36 anomalous user sessions. To further increase the interpretability of this framework, this study incorporates a voting classifier based on the use of Shapley Additive Explanation (SHAP), which identifies the most influential behavioral features used to make a prediction. This study shows that a combination of engineered behavioral features, anomaly detection through Isolation Forest, and ensemble voting can provide a lightweight, scalable, and deployable framework for real-time cloud security monitoring without requiring any intrusive biometric techniques.

Analysis of Strategies to Improve the Financial Feasibility of the Rantauprapat–Kisaran Toll Road Section

2026-09-17T06:42:18+00:00

The Rantauprapat–Kisaran Toll Road is a strategic part of the Trans-Sumatra corridor that supports connectivity and regional economic development in North Sumatra, Indonesia. However, the baseline financial analysis indicates that the project is not financially feasible, with a Financial Internal Rate of Return (FIRR) of 8.351%, a negative Net Present Value (NPV) of IDR 3,971 billion, and a Payback Period (PP) of 15.95 years. This study aims to identify appropriate strategies for improving project feasibility and assess the sensitivity of financial indicators to key parameters. A quantitative descriptive case study was conducted using Discounted Cash Flow modeling. Financial feasibility was evaluated through NPV, FIRR, and PP, while alternative strategies were assessed using single and combined scenarios involving tariff adjustments, CAPEX and O&M efficiencies, financing optimization, and Viability Gap Funding (VGF). The findings show that no single strategy satisfies all feasibility criteria. Achieving the minimum FIRR would require a tariff increase of 123.45%, VGF support of 62.84% of project costs, or CAPEX and O&M reductions of 56.91%. The optimal combination consists of 10% VGF support, CAPEX and O&M efficiencies, and a financing structure of 70% debt and 30% equity. This scenario produces an FIRR of 11.72%, a WACC of 7.56%, an NPV of IDR 12,658 billion, and a PP of 11.84 years. Sensitivity analysis identifies revenue and CAPEX as the most critical variables, highlighting the importance of revenue assurance and effective cost control.

Analysis of Mechanical Characteristics of Plain and Deep Reinforced ...Ø10 mm Through Tensile Test and Bend Test

2026-09-17T06:43:39+00:00

Reinforcing steel is the main component in reinforced concrete structures that functions to withstand tensile forces, control cracks, and ensure structural ductility. Plain rebar and 10 mm diameter deformed rebar are the most widely used in residential construction in Indonesia, but studies comparing them through two testing methods are still limited. This study aims to compare the mechanical characteristics of plain rebar Ø10 mm and D10 mm deformed rebar based on tensile and bending tests as a basis for selecting residential construction materials in Jakarta. The method used is quantitative with a laboratory experimental approach. Twelve specimens were tested at the Civil Engineering Laboratory of Esa Unggul University, each with three tensile test specimens and three bending tests per type of reinforcement, referring to ASTM E8/E8M, ASTM E290, and SNI 2052:2017. The tensile test results showed that plain rebar had an average breaking strength of 516.33 MPa and a yield strength of 375.33 MPa, while the deformed rebar had 511.67 MPa and 360.00 MPa. In the bending test, the bending strength of both was almost equal (848.33 MPa and 848.00 MPa), but the energy absorption of the deformed rebar was higher (43.15 Joules compared to 41.38 Joules) and all specimens did not crack or break. The shape of the reinforcement surface was proven to have no significant effect on strength, but rather on the ability of plastic deformation, so that deformed rebar is still recommended as the main reinforcement.

Performance Analysis of Property Construction Vendors Using the Analytical Hierarchy Process and Decision Tree Methods

2026-09-22T07:37:06+00:00

Vendor performance evaluation in property construction projects is often dominated by subjective assessments, which may lead to inconsistent decision-making. This study aims to analyse the performance of property construction vendors at PT PELNI (Persero) using the Analytical Hierarchy Process (AHP) and the C4.5 Decision Tree method. Evaluation criteria and sub-criteria were identified through expert interviews, after which priority weights were calculated using AHP. The C4.5 Decision Tree algorithm was then employed to classify vendor performance based on historical data. To address class imbalance in the training dataset, the Synthetic Minority Oversampling Technique (SMOTE) was applied. The results identified four evaluation criteria and ten sub-criteria, with Quality emerging as the most influential criterion, receiving the highest priority weight of 0.432. The C4.5 Decision Tree model achieved an accuracy of 88.89% and a recall of 83.33%, generating seven interpretable classification rules. A comparison of the two methods indicates that AHP is effective for determining the priority weights of evaluation indicators and ranking vendors, while the Decision Tree method effectively identifies classification patterns from historical performance data. The integration of these methods provides a more objective and consistent vendor evaluation framework, supporting informed decision-making in vendor selection and performance assessment at PT PELNI (Persero).

Design and Implementation of a Multi-Zone Automatic Irrigation Control System Based on Sugeno Fuzzy Logic for a Melon Seedling Incubator

2026-09-24T08:02:47+00:00

Maintaining appropriate moisture conditions is important for the early growth of melon seedlings. This study developed a zone-based irrigation control system using six capacitive moisture sensors, an Arduino Mega 2560, one water pump, and three solenoid valves. The six seedling trays were divided into three irrigation zones, with the lower value of two sensors used to represent each zone condition. Each sensor was calibrated individually, and the analog readings were converted into a dimensionless Normalized Moisture Index (NMI). A zero-order Sugeno fuzzy inference system was used to determine irrigation duration based on four input conditions: Very Dry, Dry, Moist, and Wet. Lower NMI values produced longer irrigation durations, while wet conditions produced no irrigation command. The developed system was able to process sensor readings, calculate zone conditions, generate irrigation-duration outputs, and control the pump and solenoid valves according to the designed logic. The system demonstrates the application of capacitive sensing and Sugeno fuzzy control for irrigation management in melon seedlings.

A Decision Support Model for Optimizing Capital Structure in the Jayapura–Wamena Trans-Papua Road Public–Private Partnership (PPP) Project: The Mamberamo–Elelim Segmen

2026-09-24T08:02:45+00:00

Infrastructure development through Public–Private Partnerships (PPP) has become a key strategy for addressing infrastructure financing constraints. Successful PPP implementation depends not only on selecting an appropriate partnership scheme but also on designing an optimal capital structure that ensures financial feasibility and resilience. This study develops a Decision Support Model (DSM) that integrates PPP scheme evaluation, capital structure formulation, financial performance assessment, and sensitivity analysis. A quantitative approach was applied through several stages, including evaluation of PPP scheme alternatives, Life Cycle Cost (LCC) analysis, identification of capital structure determinants, development of leverage-based financing alternatives, financial feasibility assessment, and sensitivity analysis. The results show that three PPP schemes—Design–Build–Finance–Maintain–Operate (DBFMO), Design–Build–Finance–Maintain (DBFM), and Operation and Maintenance (O&M)—met the criteria for further evaluation. Among nine combinations of PPP schemes and capital structures, only the DBFMO scheme with Debt-to-Equity ratios of 80:20 and 75:25 satisfied the financial feasibility criteria based on Weighted Average Cost of Capital (WACC), Internal Rate of Return (IRR), and Net Present Value (NPV). Sensitivity analysis identified Availability Payment as the most influential parameter affecting NPV, followed by LCC, while Cost of Debt had a relatively smaller effect. The proposed DSM provides a systematic framework for evaluating financing alternatives and supporting capital structure decisions in PPP infrastructure projects.

Optimization of the Western Surabaya Shipping Channel Design Based on Semi-Quantitative Risk Analysis for Navigation Risk Level Reduction

2026-09-24T08:02:43+00:00

The Western Surabaya Shipping Channel (APBS) serves as Indonesia's primary logistics gateway to Eastern Indonesia. Despite its 2014 revitalization, 27 shipping accidents occurred between 2014–2024, indicating persistent navigation safety gaps. This study aims to optimize APBS design through a semi-quantitative risk analysis approach to reduce navigation risk levels in eight critical segments. The methodology integrates hydro-oceanographic data, ship traffic, historical accidents, and expert judgment from six experts (harbor pilots, KSOP, Navigation District), using a 5×5 risk matrix with Analytical Hierarchy Process (AHP) weighting. Results show that the Channel Geometry Factor (CGF) dominates risk contribution (45%), followed by Exposure Frequency (23%), Environmental Conditions (20%), and Historical Accident Factor (12%) (CR = 0.077). Of 15 identified hazards, 80% are classified as Unacceptable. The combined optimization scenario (deepening, widening, geometry improvement, and shipwreck evacuation) reduces 91.7% of hazards to ALARP/Acceptable levels. However, minimum design standards (width 200 m, depth –14 mLWS, radius 4×LOA) are insufficient for Segment 3, leaving a residual risk score of R' = 16. Additional interventions—radius ≥1,000 m (5×LOA), deepening to –15 mLWS, and a one-way system—are required. This research concludes that risk-based APBS optimization effectively reduces navigation risk but necessitates interventions beyond conventional minimum standards. The findings provide actionable recommendations for policymakers and port authorities in enhancing maritime safety and operational efficiency.

Vehicle Trip Generation Modeling Based on Building Characteristics and Commercial Activities in Commercial and Service Areas

2026-09-25T07:09:28+00:00

This study aims to develop a vehicle trip attraction model based on building characteristics and commercial activities at Mie Gacoan Pandaan. A quantitative approach with a field survey method was employed. Data were collected through observation, traffic counting, documentation, and secondary sources, then analyzed using multiple linear regression with IBM SPSS Statistics. The resulting model is Y = -11.361 + 0.359X₁ + 0.455X₂ + 0.268X₃ + 0.266X₄ + 0.276X₅ + 0.448X₆. The results indicate that building area, parking capacity, seating capacity, number of visitors, operating hours, and number of employees have positive and significant effects on vehicle trip attraction. The coefficient of determination (R²) of 0.891 indicates that the model explains 89.1% of the variation in vehicle trip attraction. Parking capacity is the most dominant factor affecting vehicle trip attraction. The proposed model can be used as a reference for Traffic Impact Analysis and transportation planning in commercial and service areas.

Value Engineering in Temporary Cofferdam Works: The Impact of Using Type II Steel Sheet Piles on Cost and Time Efficiency in the Palu IV Bridge Reconstruction Project

2026-09-25T07:09:27+00:00

Temporary cofferdam works in post-disaster infrastructure projects require rapid, safe, and cost-effective technical decisions. This study evaluates the use of Steel Sheet Pile (SSP) Type II as an alternative to SSP Type IV in the temporary cofferdam works of the Palu IV Bridge Reconstruction Project. The study aims to examine the effects of stakeholder perception, material availability, and ease of implementation on cost efficiency; to analyze the effect of cost efficiency on work completion time; and to assess the implementation impact of SSP Type II in terms of material requirements, driving volume, technical control, and direct costs. A descriptive quantitative approach was applied through a survey of 36 project respondents and Structural Equation Modeling-Partial Least Squares (SEM-PLS). The SEM-PLS results show that stakeholder perception significantly affects cost efficiency (O=0.664; t=5.034; p<0.001), material availability significantly affects cost efficiency (O=0.912; t=2.397; p=0.017), ease of implementation significantly affects cost efficiency (O=1.151; t=3.484; p=0.001), and cost efficiency significantly affects project time (O=0.766; t=10.500; p<0.001). Technically, the 18 m SSP Type II system with HB 350x350x12x19 cross bracing met the criteria for stability, deflection, bending moment, and stress ratio based on the project design review. In the pier works, SSP Type II reduced the number of sheets from 420 to 400 units, decreased the driving length from 8,610 m to 7,200 m, and produced cost savings of IDR 11,938,147,636.30 or 64.84% compared with SSP Type IV. These findings indicate that value engineering can improve cost efficiency and provide volume-based time efficiency indications when supported by technical justification, stakeholder coordination, and effective implementation control.

Design of Monitoring Dashboard for Mining Pump with Design Thinking and Acceptance Analysis with Technology Acceptance Model

2026-10-01T07:56:26+00:00

PT SBS is a pump manufacturing company that conducts Factory Acceptance Tests (FAT) prior to field deployment. Current FAT procedures require personnel to physically enter hazardous zones generating three fundamental problems: high occupational safety risks, low data validity from human and parallax error, and limited information transparency for customers. This research designs a SCADA-based monitoring system using Design Thinking as the design methodology and the Technology Acceptance Model extended with TAM3 external variables as the technology acceptance evaluation framework. Design Thinking was applied through five stages (Empathize, Define, Ideate, Prototype, Test) using Node-RED, while acceptance evaluation utilized questionnaires from 60 PT SBS workshop employees analyzed via PLS-SEM using SmartPLS 4. Design Thinking produced a SCADA dashboard eliminating the requirement for operators to enter hazardous zones during testing. TAM evaluation confirmed all 26 indicators are valid (AVE > 0.50) and reliable (α > 0.80). The model explains 72.1% of Actual Usage variance. Attitude Toward Using is the sole direct predictor of Actual Usage (β = 0.849; f² = 0.882, large effect), fully mediated by PU (β = 0.454; p = 0.000) and PEOU (p = 0.005). Technology Self-Efficacy is the strongest predictor of PEOU among external variables (β = 0.494; p = 0.000), followed by Perceived Enjoyment (β = 0.270; p = 0.004). Perception of External Control and Technology Anxiety were not significant. This research concludes that integrating Design Thinking and TAM produces a system well-accepted across all user levels, with occupational safety motivation

as the primary driver of actual usage.

Interpreting a Multimodal Late-Fusion Model for Municipality-Level Community Bus and Demand-Responsive Bus Classification in Kyushu Area Using SHAP and Grad CAM

2026-10-02T03:45:32+00:00

Limited public transport accessibility remains a serious issue in aging and low-density areas such as Japan. Local governments have introduced Community Bus (CB) and Demand-Responsive Bus (DRB) services, but methods for explaining the regional patterns associated with these two service types remain limited. This paper examines the interpretability of a multimodal late-fusion model that integrates tabular regional indicators and Geographic Information System (GIS)-based spatial images for municipality-level CB and DRB classification. The research used municipalities in Kyushu, Japan. A multimodal late-fusion architecture was adopted as a commonly used integration strategy for image and tabular data. Model interpretation was conducted using SHAP for the tabular branch and Grad-CAM for the image branch. Performance was summarized using Macro F1-score as the main metric. The late-fusion model achieved a mean Macro F1-score of 0.5978 and a mean accuracy of 0.6797 under the non-TTA setting. TTA was evaluated as an inference-time robustness strategy and showed slightly smaller fold-wise variability, although the non-TTA setting produced higher mean predictive scores. Therefore, the subsequent interpretation analysis was based on the non-TTA results. SHAP showed that single elder household density, lowest elevation, elder rate, highest elevation, and total population were among the most influential features. Class-specific SHAP results also showed differences between correctly classified CB and DRB municipalities. Grad-CAM indicated that the model attended to meaningful internal spatial regions, although its interpretation was more qualitative than SHAP. These results show that multimodal late fusion can provide interpretable evidence for municipality-level CB and DRB classification and support more transparent regional transport analysisLimited public transport accessibility remains a serious issue in aging and low-density areas such as Japan. Local governments have introduced Community Bus (CB) and Demand-Responsive Bus (DRB) services, but methods for explaining the regional patterns associated with these two service types remain limited. This paper examines the interpretability of a multimodal late-fusion model that integrates tabular regional indicators and Geographic Information System (GIS)-based spatial images for municipality-level CB and DRB classification. The research used municipalities in Kyushu, Japan. A multimodal late-fusion architecture was adopted as a commonly used integration strategy for image and tabular data. Model interpretation was conducted using SHAP for the tabular branch and Grad-CAM for the image branch. Performance was summarized using Macro F1-score as the main metric. The late-fusion model achieved a mean Macro F1-score of 0.5978 and a mean accuracy of 0.6797 under the non-TTA setting. TTA was evaluated as an inference-time robustness strategy and showed slightly smaller fold-wise variability, although the non-TTA setting produced higher mean predictive scores. Therefore, the subsequent interpretation analysis was based on the non-TTA results. SHAP showed that single elder household density, lowest elevation, elder rate, highest elevation, and total population were among the most influential features. Class-specific SHAP results also showed differences between correctly classified CB and DRB municipalities. Grad-CAM indicated that the model attended to meaningful internal spatial regions, although its interpretation was more qualitative than SHAP. These results show that multimodal late fusion can provide interpretable evidence for municipality-level CB and DRB classification and support more transparent regional transport analysis.

Comparing LDA and BERTopic for Topic Modeling of Skin Complaints in Indonesian-Language Sunscreen Reviews on Shopee

2026-10-05T06:59:11+00:00

Sunscreen use in Indonesia continues to grow alongside the high ultraviolet exposure typical of a tropical climate. However, various complaints such as a sticky texture, white cast, and the appearance of acne are still frequently found in user reviews. These reviews contain important information about users' experiences and skin problems, yet they are difficult to analyze manually because of their large volume and unstructured nature. This study therefore aims to identify skin-complaint patterns in sunscreen use through an analysis of user reviews on Shopee Indonesia. It applies a text mining approach using topic modeling methods, namely Latent Dirichlet Allocation (LDA) and BERTopic. Sunscreen product reviews were collected from the Shopee platform and then passed through preprocessing stages comprising cleaning, case folding, tokenizing, normalization, and stemming. The two methods were subsequently used to extract topics representing user complaints. Evaluation was carried out using the coherence score, topic diversity, and manual interpretation to assess the quality of the resulting topics. Based on the results, LDA produced nine topics with a coherence score of 0.35 and a topic diversity of 0.70, whereas BERTopic produced twelve topics with a coherence score of 0.50 and a topic diversity of 0.90. Based on the evaluation and an examination of dominant words and representative reviews, the topics produced by BERTopic showed better interpretability because they were more specific and easier to label, whereas LDA produced more general topics that still combined several complaints within a single topic. Accordingly,

Integrating Demand Forecasting and Inventory Policy for Home Textile Products under Limited Historical Data: Evidence from a Bed Linen Manufacturer

2026-10-06T05:09:00+00:00

Demand uncertainty remains a critical challenge for textile manufacturers, particularly for bed linen products that vary in size, design, consumer preference, promotional cycles, and seasonal demand patterns. In practice, inaccurate demand planning may lead to excessive finished goods inventory or, conversely, stockouts that reduce sales opportunities and customer service performance. This study aims to develop an integrated model that links demand forecasting with inventory policy for bed linen products at PT XYZ. The study uses weekly sales and inventory data for seven bed linen size variants from August 2024 to June 2025. The research process includes exploratory demand analysis, demand pattern classification, time series forecasting model development, forecast accuracy evaluation, and the integration of forecasting outputs into inventory control decisions through safety stock and reorder point calculations. Several classical time series forecasting methods, including Moving Average, Exponential Smoothing, Holt Winters, and ARIMA/SARIMA, are compared using MAE, MAPE, and RMSE. Due to the limited historical dataset, rolling origin cross validation is applied to obtain a more reliable model evaluation. The proposed framework is expected to identify the most suitable forecasting method for each demand pattern and provide a more structured basis for inventory decision making. By integrating forecasting results into inventory policy, this study is expected to help reduce overstock and stockout risks, improve service level, and support more adaptive production planning in the textile manufacturing context.

The Impact of Overlapping Phases in EPCC Projects on the Accuracy of 4D BIM (Time) and 5D BIM (Cost): A Literature Review

2026-10-09T06:07:25+00:00

Engineering, Procurement, Construction, and Commissioning (EPCC) projects often apply overlapping phases to accelerate project completion. This strategy allows procurement, construction, or commissioning to start before the preceding phase is fully completed. However, design, material, schedule, and cost information may continue to change and affect the accuracy of 4D and 5D Building Information Modeling (BIM). This study explains the effect of EPCC phase overlap on the time accuracy of 4D BIM and the cost accuracy of 5D BIM and describes the role of BIM in project control. A literature review was conducted using 38 journal articles covering fast-track construction, 4D BIM, 5D BIM, scheduling, procurement, quantity take-off, interoperability, and cost control. The findings show that overlap can reduce BIM accuracy when design changes, procurement status, field progress, and cost information are not updated promptly. The accuracy of 4D BIM is affected in work sequence, activity relationships, duration, progress, and completion time. The accuracy of 5D BIM is affected in quantities, prices, rework costs, scope changes, and final cost forecasts. Nevertheless, BIM remains important because it enables project teams to identify the time and cost consequences of changes more quickly. BIM accuracy can be maintained through regular model updates, consistent project data, integration of models with schedule and cost information, and coordination among EPCC phases.

Comparative Analysis of Geotextile-Reinforced Embankment Slope Stability Using the Bishop Method in GeoStudio SLOPE/W and Finite Element Analysis in PLAXIS 2D: A Case Study of the Klaten–Purwomartani Toll Road Project, Section 1, Package 1.2, STA 38+000

2026-10-09T06:07:23+00:00

This study aims to analyze the stability of embankment slopes in the Klaten–Purwomartani Toll Road Construction Project Section 1Package 1.2 STA 38+000 by comparing the analysis results using GeoStudio SLOPE/W and PLAXIS 2D, as well as evaluating the effectiveness of geotextile reinforcement and verification using the Bishop method. The research data were obtained from independent laboratory testing of subgrade soil at depths of 1–3 m. The classification results indicate that the subgrade soil is dominated by clay and silt with different shear strength characteristics at each depth, while the embankment material is classified as loamy sand. The analysis without reinforcement shows that increasing embankment height and variations in loading conditions reduce the Safety Factor (SF) value, with the most critical condition occurring at an 8 m embankment height, which does not meet the requirements of SNI 8460:2017 (SF ≥ 1.5). The application of Woven Staple Fiber Polypropylene UNW-60P geotextile reinforcement effectively increases the SF value by approximately 80%, allowing all critical conditions to meet the stability requirements. Verification using the Bishop method produces SF values ranging from 1.626–2.884, all of which satisfy the requirements of SNI 8460:2017. The comparative analysis results show that PLAXIS 2D produces more conservative SF values because it considers stress distribution, deformation, and pore water pressure using the Finite Element Method (FEM) with the Strength Reduction Method (SRM), while GeoStudio SLOPE/W based on the Limit Equilibrium Method (LEM) produces relatively higher SF values and is suitable for preliminary slope stability

evaluation.

Evaluation of the Tangibles Dimension of Digital Quality and Occupational Health and Safety Application Service Quality in XYZ Construction Company

2026-10-14T08:22:01+00:00

This study evaluates the service quality of a Digital Quality and Occupational Health and Safety (OHS) application based on the Tangibles dimension of the SERVQUAL model, focusing on user interface design, ease of navigation, and the availability of supporting facilities in construction projects at Company XYZ. The application supports quality control and OHS management, making its usability essential for improving operational effectiveness. A quantitative evaluative approach was employed, involving 109 application users selected through purposive sampling. Data were collected using questionnaires and analyzed through SERVQUAL gap analysis to assess discrepancies between user expectations and perceived performance. One-way ANOVA was also conducted to identify differences in perceptions across user groups. The findings revealed negative SERVQUAL gap scores across the Tangibles dimension, indicating that the application's perceived performance had not fully met user expectations. The largest gap concerned the availability of supporting facilities, followed by ease of navigation. Furthermore, ANOVA results demonstrated significant differences in perceived performance among user groups, while their expectations showed no statistically significant differences. These findings highlight the need for improvements in the application's tangible service attributes, particularly user interface design, navigation simplicity, and the availability of comprehensive user guides and supporting resources. The study provides practical recommendations for Company XYZ to enhance user experience, improve application accessibility, and strengthen the effectiveness of digital quality and OHS management implementation in construction projects.

Investment Feasibility Analysis of the Palembang–Indralaya Toll Road

2026-10-14T08:21:59+00:00

The Palembang–Simpang Indralaya Toll Road, spanning 21.93 km, is a strategic component of the Trans-Sumatra Toll Road network, intended to improve regional connectivity and reduce logistics costs in South Sumatra. However, average daily traffic between 2017 and 2024 reached only 26%–45% of projected levels, raising concerns about the project's financial viability. This study evaluates the financial feasibility of the existing financing structure and compares alternative schemes, including the tariff-based Base Case, Availability Payment (AP), 30% Viability Gap Funding (VGF), and asset divestment at 49%, 51%, and 100%. Using a capital budgeting approach, financial performance was assessed through Net Present Value (NPV), Internal Rate of Return (IRR), and Payback Period (PP). Under the tariff scheme, the project generates an IRR of 10.98%, an NPV of IDR 1.37 trillion, and a payback period of 16 years. The 30% VGF scheme provides the strongest financial results, increasing the IRR to 12.41% and NPV to IDR 2.01 trillion while reducing the payback period to 14 years. Conversely, the AP scheme records an IRR of 4.17% and a negative NPV of IDR 635.43 billion, indicating financial infeasibility. Asset divestment also weakens financial performance as the divested share increases. These findings identify 30% VGF as the preferred financing alternative for improving investment feasibility and addressing risks arising from lower-than-expected traffic volumes. The results highlight the importance of targeted public financial support in sustaining infrastructure investment under persistent demand uncertainty. This assessment offers practical guidance for policymakers and investors selecting financing arrangements for comparable toll road developments.