Given the significance of sustainability in the current global market, doing a comprehensive and current assessment of a supplier’s sustainability might be beneficial for improving a firm’s competitiveness. This study aims to develop sustainable supplier selection criteria, examine their interrelationships, and determine the critical criteria in the paint and coating industry. This study employed a focus group discussion (FGD), the Delphi method, a decision-making trial and evaluation laboratory-based analytical network process (DANP), and a measurement of alternatives and ranking according to compromise solution (MARCOS). The findings indicated that most sub–criteria experts agreed that 11 economic sub–criteria, five environmental sub–criteria, and three social sub–criteria should be considered when selecting a sustainable supplier. The results show that the most influential criteria are environmental criteria, while the critical criteria are social criteria. In addition, “Service” (E4), “Green Material” (L3), and “Health and Safety” (S1) sub–criteria should be considered to assess the supplier because they have high relation and prominence. Globally, the top five highest weighted sub–criteria are “Respect for the Law” (S2), “Health and Safety” (S1), “Worker’s Social Rights” (S3), “Waste Management” (L5), and “Environment Management System” (L1). We practically selected 12 providers from the paint and coating sector to assess the scheme’s practicality. This work introduced a novel model for selecting a sustainable supplier that helps industries understand their core criteria and sub–criteria. This will enhance the industry’s competitiveness and sustainability by evaluating suppliers based on economic, environmental, and social criteria.

This study focuses on demand forecasting for packaging products, specifically drums and cans, employing the Random Forest Regressor (RFR) as a data-driven approach to predict packaging demand. It investigates the impact of data split scenarios and hyperparameter tuning on the forecasting testing data split ratio that produces optimal model performance. The study aims to identify the best Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) values achieved through optimized hyperparameter configurations, and to evaluate the forecasting capability of the Random Forest model in supporting operational decisionmaking for packaging inventory management. The methodology involves an experimental design with multiple data split scenarios and implements systematic hyperparameter tuning, involving the number of trees, maximum depth, and minimum samples per split. The results show that the 80:20 data split gave the best balance between learning and generalization, which is in line with previous findings in the literature. After hyperparameter optimization, the Random Forest model achieved a MAPE of $\mathbf{8. 4 4 \%}$ and an RMSE of $\mathbf{1, 9 5 9}$. This shows a significant improvement in forecast accuracy compared to the baseline configurations. This study contributes valuable insights into implementing Random Forest Regressor techniques for demand forecasting in the packaging industries, highlighting the importance of a comprehensive data split for training and validation, as well as hyperparameter tuning evaluation, to enhance forecast quality and inform operational decisionmaking in inventory management.

In the process of receiving Avtur/Jet A-1 by PT. XYZ, the product received often does not meet the specification standard on the Electrical Conductivity. To overcome, doping Stadis 450 was carried out which could increase the Conductivity Unit. In the doping process, there are inefficiencies in the use of the Stadis 450. The result obtained indicates that all the final CUs are not following the CU target with an average difference of 41.67 pS/m. Also, the current pandemic conditions have resulted in the stockpiling of Avtur/Jet A-1 in the stockpile tank for a longer time. So that PT. XYZ increases the CU target in the doping process so the CU value can be appropriate when used. Determination of CU targets in trial and error without standard considerations will lead to waste in the use of Stadis 450 and increase the potential for discrepancies that occur in the doping process. To minimize, a reduction in CU Avtur/Jet A-1 is calculated as one of the suggestions in considering the CU target to be achieved. The result of the observed decline based on hoarding data showed that there was an average decline of 6.4646 pS/m per day.