Estimating Municipal Solid Waste Using Linear Regression Modelling at MRF South Jakarta

Accurate estimation of municipal solid waste (MSW) generation is critical for effective urban waste management planning and infrastructure optimization. However, many community-based waste management facilities, such as Siaga Municipal Recovery Facility (MRF) in South Jakarta, face challenges in direct waste quantification due to the common absence of weighbridge systems, leading to a gap in precise data for operational and strategic decision-making. This study addresses this gap by developing and applying a simple linear regression modelling approach to estimate MSW generation. The model utilizes readily available volume data collected from various types of solid waste transport units operating at Siaga MRF. A dataset comprising 46 observational data points, including actual solid waste weight and corresponding volume measurements, was collected over several operational days. The regression model was developed and validated using a data mining approach with Python, employing the Scikit-Learn and Pandas libraries to identify the statistical relationship between conveyance volume (input) and waste weight (output). The linear regression model yielded an R-squared (R²) value of 0.562, indicating a moderate but practically useful level of predictive accuracy in explaining the variance in waste weight based on volume. The Mean Squared Error (MSE) was 0.0125, further supporting the model's reliability for estimation purposes in this context. These findings demonstrate that simple linear regression can serve as a feasible, low-cost, and valuable tool for estimating waste generation where direct weighing infrastructure is unavailable. The results are intended to assist policymakers, facility operators, and urban planners in enhancing capacity planning, improving daily operational forecasting, and guiding future investments in waste management technologies. Furthermore, this study reinforces the importance of applying quantitative, data-driven approaches to promote evidence-based environmental management and operational efficiency at the community scale.