Developing a Robust Optimisation Framework for Production Planning Under Demand and Supply Uncertainty in Nigerian Consumer Goods Manufacturing

This dissertation develops a robust optimisation framework for production planning under simultaneous demand and supply uncertainty in Nigerian consumer goods manufacturing, addressing the theoretical and methodological gap between production planning models developed under certainty assumptions and the high-uncertainty operating environment faced by Nigerian manufacturers. Nigerian consumer goods manufacturers contend with demand uncertainty arising from volatile disposable income dynamics, seasonal consumption patterns, and competitive disruptions, alongside supply uncertainty from raw material import delays, domestic agricultural supply variability, and energy availability unpredictability. The research formulates an original two-stage robust stochastic programming model that minimises the worst-case production cost across a defined uncertainty set while satisfying service level constraints. The uncertainty set is characterised empirically from historical demand and supply data collected from seven consumer goods manufacturers in Lagos and Ogun states, and its structure is analysed to inform the design of a computationally tractable robust counterpart formulation. A decomposition algorithm exploiting the structure of the two-stage problem is developed to enable solution at commercially relevant problem scales. The framework is implemented as a decision support system integrated with the company's SAP enterprise resource planning environment at one pilot manufacturer. Experimental results demonstrate mean cost savings of twelve-point-three percent against the company's existing deterministic planning approach while achieving service level improvements of nine-point-eight percentage points. Keywords: robust optimisation, production planning, demand uncertainty, stochastic programming, Nigerian consumer goods