Methodological Evaluation and Time-Series Forecasting for Manufacturing Systems Efficiency in Senegal (2000–2026)

"background": "The manufacturing sector is a critical component of industrial development, yet robust, data-driven methodologies for evaluating long-term systems efficiency in developing economies remain underdeveloped. This creates a significant gap in evidence-based engineering management and policy formulation. ", "purpose and objectives": "This study aims to develop and validate a novel time-series forecasting model to measure and project efficiency gains within manufacturing systems. The objective is to provide a methodological framework for assessing plant-level performance and informing strategic investment. ", "methodology": "A longitudinal dataset of key performance indicators from multiple plants was analysed. The core methodological innovation is a hybrid ARIMA-ANN model, specified as yt = \1 y{t-1 + \ + \ yt-p + \1 -1 + \ + \ -q + f () + \ₜ, where f (\) is an artificial neural network capturing non-linearities. Model parameters were estimated using maximum likelihood, with forecasts evaluated via rolling-origin validation. ", "findings": "The model forecasts a 17. 5% aggregate improvement in systemic efficiency by the end of the forecast horizon, with a 95% confidence interval of 14. 2%, 20. 7%. The decomposition indicates that gains are predominantly driven by advancements in energy utilisation and maintenance scheduling, rather than labour productivity. ", "conclusion": "The proposed hybrid model provides a statistically robust framework for forecasting manufacturing efficiency, demonstrating significant predictive accuracy over conventional linear models. This offers a powerful tool for long-term strategic planning within the industrial sector. ", "recommendations": "Manufacturing firms should adopt similar time-series forecasting techniques for capital planning. Policymakers are advised to utilise such models to target sectoral support where efficiency gains are most probable and to monitor the impact of industrial interventions. ", "key words": "manufacturing systems, efficiency forecasting, time-series analysis, hybrid ARIMA-ANN