Evaluating stress behavior in prestressed concrete girders during staged construction is important for structural safety. This study compares stress responses based on three design codes: AASHTO, Eurocode, and IRC. A total of 21 construction stages were simulated using MIDAS Civil 2023 to model a balanced cantilever bridge. The analysis included time-dependent effects such as creep, shrinkage, and prestress losses, following each code’s assumptions. To ensure fair comparison, material properties, geometry, and environmental conditions were kept uniform. Results show that the highest compressive stress at the top fiber occurred in the Eurocode model (11,960 kPa), while the highest tensile stress was found in the IRC model (841.3 kPa). At the bottom fiber, IRC also produced the highest compressive (18,630 kPa) and tensile stresses (1,282 kPa). These differences indicate variations in how each code accounts for viscoelastic behavior and stress redistribution. This study highlights the importance of developing a national design standard in Indonesia that balances safety and efficiency by adapting insights from international codes.

The construction of toll roads requires supervision consultants capable of ensuring project quality, time, and cost performance, yet a structured and measurable assessment framework is lacking. This study develops a comprehensive performance assessment rubric for toll road supervision consultants using quantitative methods, including a literature review, a survey of 50 respondents, validation by six experts, and an application test by 12 evaluators. Data were analyzed using validity and reliability tests, Mann–Whitney U, AHP, TOPSIS, and ROC methods. The results identified 18 valid and reliable indicators grouped into three dimensions soft skills, technical skills, and project supervision. The most critical indicators were QA/QC (C=0.810), periodic reporting (0.782), and work volume management (0.711). Final dimension weights were 40% for technical skills, 30% for soft skills, and 30% for project supervision. Application of the rubric indicated consultants’ performance remained in the “Average” category, highlighting areas for measurable improvement in supervision practice.

The safety and efficiency of oil and gas pipelines remain critical due to the high risks associated with leaks, pressure surges, and undetected structural damage. WSNs provide an effective solution for real-time monitoring by deploying sensor nodes along the pipeline. However, existing routing protocols such as DSDV and AODV face challenges with congestion, packet loss, and uneven energy consumption in long linear topologies. This study proposes DSDVTRI, a triple-interleaving extension of the DSDV protocol designed to improve data delivery and energy efficiency in linear pipeline networks. Simulations were performed using NS2.35 version across node counts ranging from 20 to 200. The results show that at number of nodes 100, DSDVTRI improves throughput by 14.2% and delivery ratio by 3.05% compared to DSDV, while reducing energy consumption per packet by 6.3%. These findings demonstrate that DSDVTRI enhances performance stability, making it suitable for real-time and energy-efficient pipeline monitoring applications.

This study investigates the optimization of manufacturing femoral components for Total Knee Replacement (TKR) using Computer-Aided Manufacturing (CAM) simulation and 5-axis CNC milling, followed by dimensional verification based on 3D scanning. The machining process was simulated in Autodesk PowerMill to generate collision-free toolpaths for AISI 316L stainless steel. Dimensional verification was conducted by comparing the 3D-scanned physical model (using Creality CR-Scan Ferret Pro) with the original CAD model in Geomagic Control X. The metrological analysis showed a Root Mean Square (RMS) deviation of 0.5317 mm and an average positive deviation of 0.2572 mm. Spatial deviation analysis revealed significant dimensional variations, with a maximum deviation of +2.5761 mm and a minimum deviation of -2.5713 mm. Specifically, in critical functional regions, the medial and lateral condyles exhibited deviations ranging from -0.4683 mm to 0.232 mm, while the patellar groove showed a deviation of 0.1989 mm. Although the machining strategy successfully produced the complex implant geometry, the tolerance distribution data indicated that only 17.22% of the surface fell within the strictly specified tolerances, highlighting the need for further optimization of cutting parameters and fixturing strategies to minimize surface roughness and dimensional inaccuracies.

Catastrophic forgetting remains a major challenge in continual transfer learning, where performance on earlier tasks degrades after sequential adaptation. While full fine-tuning updates all parameters and achieves strong performance on new tasks, it is computationally expensive and prone to forgetting. This study compares parameter-efficient fine-tuning (PEFT) methods—adapters, additive learning, side-tuning, LoRA, and zero-initialized layers—against full fine-tuning on CIFAR-100 using a two-stage protocol: task-A (classes 0–49) followed by task-B (classes 50–99), evaluated on ResNet-18 and ResNet-50. Results are reported as mean ± standard deviation over three runs (n = 3), with retention measured using a Swapback-based recall method that distinguishes true forgetting (Δ). Across both architectures, all PEFT methods maintain task-A knowledge (Δ = 0.00), while full fine-tuning exhibits forgetting (Δ = 0.31 on ResNet-18; Δ = 0.20 on ResNet-50). PEFT methods achieve competitive task-B performance while updating only 0.22–4.49% of parameters. Notably, LoRA on ResNet-50 achieves the highest task-B accuracy (0.82) with only 0.93% parameter updates and no forgetting, slightly outperforming full fine-tuning (0.81). These findings highlight PEFT as an efficient and stable alternative for scalable continual transfer learning.

In order to develop molecular magnetic compounds, oligonuclear cobalt complexes with Schiff-base ligands, 1,3-bis(salicylideneamino)-2-propanol (H3salpr), 1,3-bis(5-chlorosalicylideneamino)-2-propanol (H3clsalpr), 1,3-bis(5-bromosalicylideneamino)-2-propanol (H3brsalpr), and 1,3-bis(5-nitrosalicylideneamino)-2-propanol (H3pnsalpr), were synthesized by reaction of cobalt(II) acetate and Schiff-base ligand in organic solvent. The molecular structures of [CoII{CoIII(µ-HL)(µ-CH3COO)(CH3COO)}2] (HL = Hsalpr and Hbrsalpr), [CoII{CoII(µ-Hsalpr)(µ-CH3COO)}2], and [CoIII2CoII4(µ-L)2(µ3-CH3O)2(µ-CH3O)6(CH3OH)2] (L = salpr, clsalpr, brsalpr, and pnsalpr) were elucidated by the single-crystal X-ray crystallography. The electronic spectra and magnetic susceptibilities with variable-temperature are in accordance with the trinuclear and hexanuclear molecular structures. The cyclic voltammograms are basically featureless, giving a clue to isolate reduced complexes [CoII{CoII(µ-Hsalpr)(µ-CH3COO)}2] and [CoII2(Hpnsalpr)2(CH3OH)]. These findings may guide future design of molecular magnet materials.

This study analyzes the reasons for waste in the production processes of a paper-based food packaging company in Indonesia and proposes improvement measures to make the operations more efficient. Through the application of Value Stream Mapping (VSM) and Single-Minute Exchange of Dies (SMED), three significant waste categories were charted: product defects (24.74%), unnecessary movements (22.23%), and unnecessary inventory (13.29%). Implementation of SMED successfully reduced the time of tool change from 60.24 minutes to 27.12 minutes, a setup time improvement of 54.98%. Pursuit of labeling system improvement as well as putting in place a Just-In-Time (JIT) inventory practice also played an important role in the reduction of wastes and fast responsiveness. Consequently, Process Cycle Efficiency (PCE) increased from 70.37% to 73.41%, production capacity increased by 1,241 units per cycle, and complaint counts for defective products declined appreciably over a period of six months. These results underscore the actual benefits of Lean methods and underscore the standard for performance excellence for the same pursuit of sustainable operational excellence within the same industry.

The growing need for efficient night lighting in natural tourist destinations highlights the importance of reliable and sustainable energy solutions. This study analyzes the optimization of solar-based lighting at Bukit Kunci, Indonesia, using the Open Energy Modelling Framework (OEMOF) combined with real-time monitoring via the IoT ThingSpeak platform. Photovoltaic (PV) panel data recorded at 15-second intervals during February–July 2025, yielding 532,520 records, were cleaned and aggregated as input to model the interaction of PV, batteries, LED lights, inverters, and backup generators, to minimize lifecycle cost and energy loss. Results indicate that the current PV capacity (0.4 kWp) supplies less than 50% of lighting demand, with a high levelized cost of energy (≈9.2 USD/kWh) and low reliability (self-sufficiency 3–22%). Optimization through capacity expansion (≈224 modules, ≈1.25 kWh storage) eliminated load loss probability and reduced LCOE to ≈0.05 USD/kWh. This approach demonstrates OEMOF’s potential to enhance system efficiency, ensure reliable night lighting, and support eco-tourism while offering replicability for rural destinations.

This study aims to assess the potential for urban inundation and formulate flood mitigation strategies based on spatial analysis. A 2D modelling approach using HEC-HMS and HEC-RAS was applied to simulate the extent and depth of inundation in the urban area of Karanganyar, Indonesia. The models were validated with field observations using RMSE. Model validation confirmed the agreement between simulated and observed data. Validation was quantified using RMSE = 0.42 m across 15 checkpoints, with an average MAE of 0.31 m, and a coefficient of determination (R²) of 0.87. The 2D mesh resolution was set at 10 m, balancing computational efficiency with spatial accuracy. The maximum inundated area reached 3.8 million m² (±0.2 million m²), and the 95th percentile inundation depth was 4.7 m, concentrated in residential and agricultural zones. The results were used to delineate risk zones for prioritised flood mitigation planning. The simulation identified risk zones for prioritised flood mitigation planning, revealing inundated areas of up to 3.8 million m² with maximum depths around 4.7 m (±0.3 m), primarily affecting residential areas, agriculture, and public areas. Risk-based zoning was used to prioritise mitigation strategies, including improved drainage and the construction of infiltration ponds. The risk-based zoning approach demonstrated that implementing infiltration ponds and improved drainage in Priority Zone A could reduce flood exposure by approximately 28% under a simulated 20-year return period storm. These findings provide a measurable basis for adaptive flood mitigation strategies in urban areas.

In this study, we systematically analyzed all isomers of 5–11 atom clusters using a graph-theoretical approach and evaluated their maximum Hückel energy (HE) within the Hückel approximation. The maximum HE increased with the number of bonds, reached a maximum at an intermediate value, and then decreased again, with the peak located near the average of the minimum and maximum possible edge numbers. This relationship was well reproduced by a simple linear expression. The isomers with maximum HE were found to have nearly uniform degree distributions, and classification by graph radius (Gr) and maximum degree (Ma) revealed that they possess compact and homogeneous structural characteristics. By extrapolating these trends, structural features of maximum-HE isomers were predicted for clusters with 12–14 atoms. These results demonstrate the effectiveness of a graph-theoretical description for analyzing cluster stability and provide a foundation for extending the analysis to larger clusters and for comparison with first-principles calculations.