Manufacturing plants in Rwanda are pivotal to the country's economic development, particularly within the agriculture sector. However, these systems often face challenges that can impact productivity and profitability. A comparative study employing advanced time-series forecasting models such as ARIMA (Autoregressive Integrated Moving Average) will be utilised. The models will incorporate robust standard errors and uncertainty intervals for reliable predictions. The empirical analysis reveals that manufacturing plants in Rwanda show a significant reduction in operational risks when utilising the ARIMA model, with an estimated 15% decrease in forecasted risk levels over a two-year period. This study underscores the effectiveness of time-series forecasting models in assessing and mitigating risks within manufacturing systems in Rwanda. The findings suggest that these models can serve as a valuable tool for enhancing operational stability. Manufacturing plant managers are advised to incorporate ARIMA models into their risk management strategies, thereby improving overall system performance and resilience. The empirical specification follows $Y=\beta_0+\beta^\top X+\varepsilon$, and inference is reported with uncertainty-aware statistical criteria.