Methodological Evaluation and Time-Series Forecasting for Yield Improvement in Senegalese Process-Control Systems

{ "background": "Process-control systems in industrial settings are critical for operational efficiency and yield optimisation. In many developing economies, including those in West Africa, there is a recognised need to move beyond basic system implementation towards advanced, data-driven methodologies for continuous improvement. However, methodological frameworks for evaluating these systems and forecasting performance gains are often not tailored to local operational constraints and data availability.", "purpose and objectives": "This study aims to develop and validate a methodological framework for evaluating process-control systems, with the core objective of creating a robust time-series forecasting model to predict and quantify yield improvement. The work seeks to provide a practical tool for engineers to measure the efficacy of system interventions.", "methodology": "A hybrid modelling approach was employed, integrating system dynamics analysis with statistical forecasting. The core forecasting model is an autoregressive integrated moving average with exogenous variables (ARIMAX), specified as Yt = \μ + \∑i=1ᵖ\ Yt-i + \ + \∑i=1q\ \εt-i + \∑j=1ʳ\ Xj,t. Model parameters were estimated using maximum likelihood, and robust standard errors were calculated to account for heteroscedasticity. The methodology was applied to a longitudinal dataset from a Senegalese industrial plant.", "findings": "The ARIMAX model demonstrated significant predictive capability, with a key process parameter (raw material consistency) showing a positive coefficient of 0.45 (95% CI: 0.38, 0.52). The forecasting analysis indicated that targeted system adjustments could yield a mean improvement in process efficiency of approximately 7.3% over a standard operational cycle.", "conclusion": "The proposed methodological framework and forecasting model provide a technically sound and context-adapted approach for quantifying the impact of process-control interventions. This moves yield management from a reactive to a predictive practice.", "recommendations": "Industrial engineers should adopt structured methodological evaluations