Abstract:
Inventory management is a crucial aspect of humanitarian operations. Various inventory models and policies have been developed over the years to improve the efficiency of humanitarian inventory management. These models consider various elements, including sourcing, storage, prepositioning, distribution, and transportation. While the existence of literature and models supplied guidance and breakthroughs towards more informed decision-making, the complex setting of disasters has continued to preclude their application. Over-simplification, impracticality, and particularity of decision variables pose a challenge in using specific models in exceptionally distinct disasters owing to their complexity and ever-changing nature. This implies that the ability to manage inventory efficiently and its distribution depends on the preparedness and prevailing conditions in the post disaster period.
This study focused on approaching these shortcomings by adopting an integrated approach which starts with the characterisation of inventory management challenges unique to disaster settings. Gaps within developed models are identified, and an inventory prepositioning and aid distribution model is developed and applied to bridge some gaps.
Therefore, this study presents two models (deterministic and stochastic programming with recourse) for prepositioning modelling. The models are implemented as multi-objective mixed-integer linear programming relief inventory prepositioning models for the Democratic Republic of Congo (DRC) and Central African Republic (CAR). The models minimise shortages and enhance equitability while minimising the total response time in areas with poor road network in a cross-border distribution setting. The model is solved using a pre-emptive optimisation approach, and a sensitivity analysis is conducted to evaluate the influence of the budget, priority items proportion, and capacity variation in the model input.
Results indicate that the models are sensitive to changing parameters. Of the two models, the stochastic model was determined to have higher reliability but required a higher budget to match the performance of the deterministic model. Results analyses confirm that the models can add value to humanitarian organisations when planning facility locations, inventory prepositioning, and conflict area-distribution centre assignments in the DRC and CAR. This study, therefore, contributes to the body of knowledge and humanitarian organisations in Africa.
