Transportation is critical, especially in rural areas as it provides the mobility to people to access different activities satisfying their daily needs. The purpose of this research is to create an artificial neural network (ANN) trip generation model (TGM). 500 households (HH) were surveyed to obtain the independent variables used in the modeling process including the HH size (HHS), number of children in the HH below 7 years old (NCHHBS), number of HH member from 7 to 59 years old (NHHMSF), number of HH member above 59 years old (NHHAF), number of working member of the HH (NWMHH), number of school children in the HH (NSCHH), number of helpers in the HH (NHHH), number of motorize vehicles in the HH (NMVHH), HH income (HHI), highest educational attainment (HEA), head of HH age (HHHA), and number of driver’s license holder in the HH (NDLHHH). Using the Levenberg–Marquardt algorithm (LMA) as the training algorithm (TA) and hyperbolic tangent sigmoid (HTS) function as the activation function (AF), the governing TGM was observed in the 12–25-1 network structure with the highest R value = 0.98476 and least Mean Absolute Percentage Error (MAPE) of 9.04%. Moreover, the governing network structure achieved the minimized Akaike Information Criterion (AIC) value at 25 hidden neurons (HN) indicating that the network has already been generalized and was the best model among those observed in this study. The outcomes of the research showed the efficacy of artificial neural networks in developing trip generation prediction models (PM).

The COVID-19 pandemic disrupted work systems, family, and social life. The mandatory lockdown forced employees to shift to work-from-home (WFH) setup which exposes them to WFH conflicts. This study provides a machine learning—based approach for prioritization of factors affecting WFH conflicts during the COVID-19 pandemic. These factors include time spent with the family (F1), leisure activities (F2), household task (F3), family quality of life (F4), agitation and anger from work (F5), financial obligations (F6), family presence (F7), family issues (F8), health-related (F9), and work burn-out (F10). Using the backpropagation (BP)-artificial neural network (ANN) modeling and Olden’s connection weights (CW) approach, the order of influence of these parameters to the productivity rating (PR) was observed. Based on the results, the 10-21-1 network structure is the best performing model (BPM) with correlation coefficient (R) = 0.98173 and mean squared error (MSE) of 0.02607. This network topology also provided the least Akaike Information Criterion (AIC) value showing that it is the best model. Using its connection weights (CW) through Olden’s approach, the results showed that the financial obligations are the most influential parameter (MIP) while the household task is the least influential parameter to the productivity model. The utilization of machine learning techniques proved to be effective in determining the influence of predictors on the target output. The obtained findings from the study could assist the organization and managers in resolving work-from-home conflict and productivity issues.

Sustainable development on remote islands encounters hurdles such as environmental, logistical, and economic constraints. Small islands face insufficiency and limitations in construction materials due to limited natural resources. Lightweight structures emerge as transformative solutions to overcome challenges in construction. Santiago Island faces construction complications intensified by adverse weather and logistical issues, particularly disruptions in material transport. This study proposed a lightweight structure by addressing logistical challenges, environmental challenges, and structural costs, assessing existing materials including bamboo, exploring sustainable design principles, incorporating structural standards, and constructing and evaluating a typhoon-resistant structural system. The study results of the identification of lightweight construction materials and techniques available on Santiago Island and the investigation of a sustainable design system, which is the A-frame house design or triangular-shaped structure, foster sustainable development and resilience for remote islands. Structural analysis of the lightweight sustainable design was also identified through STAAD to comprehensively check all the structural members. The results of the research propose a hopeful remedy for crafting a cost-effective, environmentally friendly, and durable architectural blueprint tailored specifically for communities residing on small islands within low- to middle-income nations.

The factor evaluation for sustainability implementation in post-disaster reconstruction projects was conducted using artificial neural network (ANN)-based sensitivity analysis. The current study involves the utilization of Levenberg-Marquardt algorithm (LMA) for the ANN model development. The findings demonstrated that the governing network structure is 10-3-1 utilizing input parameters (IP) including waste management (Sl), environmental impact management (S2), support of management (S3), construction cost (S4), public health and safety (S5), user security (S6), noise pollution (S7), energy consumption (S8), public services (S9), and recycling (S10). The results revealed that the governing model (GM) has a performance of correlation (R) = 0.99813 and Mean Squared Error (MSE) and Mean Absolute Percentage Error (MAPE) of 0.00110 and 0.4240/0, respectively. Moreover, a sensitivity analysis (SA) using Garson's algorithm (GA) reveals a trend of the relative value (RV) was observed to be construction cost > public services > support of management > user security > environmental impact management> noise pollution> public health and safety> energy consumption > recycling > waste management wherein the construction cost is the most influential parameter to the sustainability rating in post-disaster reconstruction projects (PDRP). The study shows how well ANN can detect the crucial elements that determine a project's sustainability rating in the post-disaster reconstruction projects.

In recent years, the construction sector in the Philippines has faced significant challenges stemming from various events and occurrences, leading to cost overruns and delays in project timelines. A critical element for every construction undertaking's accomplishment is cost evaluation. Precisely approximating the cost of a project involves thorough consideration of various elements, making it a difficult undertaking to forecast. Several building constructions nowadays produce high cost overrun because of unforeseen change in the project budget that raises the overall project cost such as the complexity of the building system and the organization’s environment. The aim of this paper is to offer a potential prediction for cost estimation, with the goal of minimizing the substantial risk of cost overruns in low- to mid-rise buildings. In this study, the structural elements for low- to mid-rise buildings were utilized from building constructions, such as the number of exterior walls (QEW), type of construction material (TCM), building height (HB), total gross area (TGA), building footprint area (BFA), type of occupancy (TO), number of floors (NF), quantity of shear walls (QSW), and number of columns (NC); an artificial neural network (ANN) model was employed in this research to establish a model for forecasting the total construction cost (TCC). With a correlation value (R) of 0.999890 and a mean absolute percentage error (MAPE) of 0.601%, the modeling results shown that the best model structure was 9-25-1 (input-hidden-output), indicating its effectiveness and efficacy in forecasting the TCC. The impact of each variable employed as an input variable (IV) in the model establishment was seen employing the connection weights (CW) through Garson’s algorithm (GA). The calculation exhibited the order of influence observed as QSW > NC > HB > NF > QEW > TGA > BFA > TO > TCM, wherein the quantity of the shear walls is seen to have the most contribution to the construction cost. Moreover, to check its performance versus other prediction modeling tools, a multiple linear regression (MLR) model was also created and compared to the governing prediction model (GPM). The MAPE of the BP-NN is 7.108 times better than that of the created MLR model.

Delays in bridge construction are crucial problems that slow down the economic development in an area. In this study, an artificial neural network (ANN) model was utilized to create a model for predicting the duration delay in bridge construction projects which includes the project amount, length of bank protection, length of bridge approach slope protection, total area of bridge approach, number of item of works, number of foundations, type of foundation, number of girders, type of girder, number of lanes, number of spans, total width, total length, and type of construction as the independent variables (IV). The modeling results showed that the best performing model is the 14–14–1 network with R = 0.99406 and MAPE of 3.524%. By removing each of the parameters, the influence of the independent variables to the duration delay was determined. Using the sensitivity index method, the findings revealed that the ranking of influence of the factors (IF) to the duration delay was observed as LBASP > NS > TW > TC > NIW > LBP > PA > TABA > TL > NG > NF > NL > TG > TF with the length of bridge approach slope protection was seen to be the most influential parameter (MIP) to the duration delay.

Corrosion stands as the primary cause behind the diminished service life of reinforced concrete structures, particularly in environments such as ports, harbors, bridges, and other offshore and near-shore locations where chloride-induced corrosion poses a significant threat. This study investigates the efficacy of three key parameters polypropylene fiber ratio (FR), concrete cover (CC), and bar diameter (∅)—in minimizing corrosion (CL) in reinforced concrete structures. Central Composite Design (CCD) of Response Surface Methodology (RSM) is employed to determine the optimal conditions for these parameters. The number of samples for Impressed Current (IC) testing is determined through this methodology. Initial analysis utilizing the full quadratic model yields a Predicted value R2 of 60.77%. However, employing backward elimination enhances the predictive capability, resulting in an improved R2 value of 87.53%. Sensitivity analysis utilizing the coded units of the RSM model reveals that the polypropylene fiber ratio exerts the most significant impact on corrosion levels, with a sensitivity value of -4.932. Consequently, optimization efforts are focused on this parameter, leading to the identification of an optimized value of 1.17% for FR, which results in minimal corrosion. This research underscores the effectiveness of employing RSM techniques in optimizing corrosion mitigation strategies in reinforced concrete structures, with FR emerging as a critical determinant in achieving corrosion resistance.

In this paper, the performance of feed-forward backpropagation (BP)—artificial neural network (ANN) was evaluated in predicting the construction project cost (CPC). The models include several factors involving the floor area (FA), number of floors (NF), structural material type of the building (MT), height of the building (HB), number of columns (NC), area of the concrete hollow blocks wall (CHB), volume of concrete (VC), weight of steel (WS), and contract duration (CD). The developed neural network model was evaluated based on several accuracy metrics such as correlation coefficient (R), mean squared error (MSE), mean absolute percentage error (MAPE), and Akaike Information Criterion (AIC). The simulation results showed that the governing model has an excellent R value of 0.99885 and a MAPE of 2.5826%. The comparison results between the ANN and multiple linear regression (MLR) suggest that the ANN model provided superior performance with MAPE which is 4.777 times better than that of the MLR model. Moreover, the lowest AIC value was observed in the 9–21-1 network structure suggesting that this is the governing network model for predicting the CPC. The sensitivity analysis (SA) using Garson’s Algorithm (GA) quantitatively determines the relative contribution (RC) of the input parameters (IP) to the construction project cost. The developed model could be utilized as a support instrument for minimizing the cost overruns and losses that may have been incurred in a construction project (CP).

Wind forces on structures have the potential to cause significant damage. A database involving the distance from the ridge, enclosure classification, surface type, elevation above ground level, wind direction, basic wind speed, presence of wall/surface openings, and effective net wind pressure (ENWP) was created using computation fluid dynamics (CFD). This paper focuses on the development of a model for predicting ENWP using a backpropagation-artificial neural network (BP-ANN). Utilizing the Levenberg-Marquardt algorithm (LMA) and hyperbolic tangent sigmoid function (HTSF) as the model hyperparameters, the study investigated several network structures and the simulations revealed that the 7-20-1 is the best model among the topologies observed in this study. The results showed an R value of 0.99868, MSE and MAPE of 0000749 and 5.036%, respectively. Additionally, the Akaike Information Criterion (AIC) was used as another layer of metric to measure the effectiveness of the model. The least was observed in the 7-20-1 network structure indicating that this is the best among the topologies observed in this study. Moreover, a sensitivity analysis (SA) through Garson's Algorithm (GA) was performed to determine the relative contribution (RC) of the input parameters (IP) including the distance from the ridge, enclosure classification, surface type, elevation above ground level, wind direction, basic wind speed, and presence of wall/surface opening to the effective net wind pressure. The findings presented that the basic wind speed is the most significant parameter to the effective net wind pressure value. The results of this study can be utilized in considering appropriate configuration to minimize the effects of wind pressure in structures.

Water is a resource that all living things require, especially humans. Water contamination is prevalent worldwide due to progress and rapid industrialization. This study aims to conceptualize a low-cost and straightforward to install solar photocatalytic reactor (PCR) prototype for the degradation of Methylene Blue (DMB), one of the contaminants released into water resources, with the aid of Biochar and Titanium Dioxide as nano catalysts and assess its efficiency. A 3mL sample was collected before the start of the experiment, another 3mL sample from an unilluminated setting, and a 3mL sample at the end of the 2-hour photocatalytic investigation. The samples were processed and observed from an external laboratory. Results showed a 90.53% adsorption efficiency rate, a 1.1% and 9.9% difference from another study that utilized the same contaminant and time duration, and a 0.0196/min degradation rate. Based on this result, it was assessed that the proposed photoreactor was solar adsorption efficient and had a photodegradation potential to reduce the Methylene Blue (MB) contaminant.