Managing inventory is tricky when demand is all over the place. This is especially true for intermittent demand, where sales can spike or drop unexpectedly. This unpredictability can lead to lost sales or excess stock, both of which are costly. The key to handling this is finding the right inventory policy.

    To tackle this issue, a smart approach using Markov processes and metaheuristic methods has been developed. The idea is to use past demand data to set optimal stock limits. This way, companies can strike a balance between holding too much stock and losing sales.

    The approach starts by modeling past intermittent demand using a Markov process. This helps in setting lower and upper stock limits. The goal is to minimize the biggest costs associated with intermittent demand: holding costs and lost sales.

    To test this approach, demand data was generated in four different sizes, from small to large. The results showed that using the Markov process significantly improved inventory management. It helped in balancing the increased costs due to intermittent demand.

    However, there was a hiccup. A mathematical model was proposed for optimizing stock levels, but it didn't provide a feasible solution. So, the model was converted into a fitness function. Then, two algorithms, Tabu Search and Simulated Annealing, were used to find a solution.

    The inventory management process was first evaluated without the Markov approach. Then, the Markov approach was included. The results were clear: the Markov approach was a valuable tool for managing intermittent demand.

    The stock limits computed with the Markov process did a great job of balancing the costs. This means companies can manage their inventory more effectively, even when demand is unpredictable.

Smart Stock Strategies for Unpredictable Demand

Actions