Abstract:Capacity planning decisions affect a significant portion of future revenue. In the semiconductor industry, they need to be made in the presence of both highly volatile demand and long capacity installation lead‐times. In contrast to traditional discrete‐time models, we present a continuous‐time stochastic programming model for multiple resource types and product families. We show how this approach can solve capacity planning problems of reasonable size and complexity with provable efficiency. This is achieved … Show more
“…With stochastic demand, Huh and Roundy (2005) assume lost sales and no inventory carry-over; hence there is no temporal dependence in their problem setup. Their approach is extended by Huh et al (2006).…”
We consider the capacity planning problem during a product transition in which demand for a new-generation product gradually replaces that for the old product. Capacity for the new product can be acquired both by purchasing new production lines and by converting existing production lines for the old product. Furthermore, in either case, the new product capacity is "retro-fitted" to be flexible, i.e., to be able to also produce the old product. This capacity planning problem arises regularly at Intel, which served as the motivating context for this research. We formulate a twoproduct capacity planning model to determine the equipment purchase and conversion schedule, considering (i) time-varying and uncertain demand, (ii) dedicated and flexible capacity, (iii) inventory and equipment costs, and (iv) a chance-constrained service level requirement. We develop a solution approach that accounts for the risk-pooling benefit of flexible capacity (a closed-loop planning approach) and compare it with a solution that is similar to Intel's current practice (an open-loop planning approach). We evaluate both approaches with a realistic but disguised example and show that the closed-loop planning solution leads to savings in both equipment and inventory costs, and matches more closely the service level targets for the two products. Our numerical experiments illuminate the cost tradeoffs between purchasing new capacity and converting old capacity, and between a level capacity plan versus a chase capacity plan.
“…With stochastic demand, Huh and Roundy (2005) assume lost sales and no inventory carry-over; hence there is no temporal dependence in their problem setup. Their approach is extended by Huh et al (2006).…”
We consider the capacity planning problem during a product transition in which demand for a new-generation product gradually replaces that for the old product. Capacity for the new product can be acquired both by purchasing new production lines and by converting existing production lines for the old product. Furthermore, in either case, the new product capacity is "retro-fitted" to be flexible, i.e., to be able to also produce the old product. This capacity planning problem arises regularly at Intel, which served as the motivating context for this research. We formulate a twoproduct capacity planning model to determine the equipment purchase and conversion schedule, considering (i) time-varying and uncertain demand, (ii) dedicated and flexible capacity, (iii) inventory and equipment costs, and (iv) a chance-constrained service level requirement. We develop a solution approach that accounts for the risk-pooling benefit of flexible capacity (a closed-loop planning approach) and compare it with a solution that is similar to Intel's current practice (an open-loop planning approach). We evaluate both approaches with a realistic but disguised example and show that the closed-loop planning solution leads to savings in both equipment and inventory costs, and matches more closely the service level targets for the two products. Our numerical experiments illuminate the cost tradeoffs between purchasing new capacity and converting old capacity, and between a level capacity plan versus a chase capacity plan.
“…This section presents a heuristic algorithm to solve the Capacity Planning Formulation presented in Section 2. Our algorithm is a modification of the divide-and-conquer method by Huh and Roundy [18].…”
Section: Solution Approachmentioning
confidence: 99%
“…We describe below each subroutine in detail. These subroutines take an advantage of the following observation by Huh and Roundy [18]. While the objective function F is not separable, it is quasi-separable, i.e., its partial derivative with respect to a variable τ m,n depends on the value of other variables only through the set consisting of all variables that are less than τ m,n .…”
Section: Solution Approachmentioning
confidence: 99%
“…The boundary is drawn by the Convexity Conditions summarized in Section 2.3. Huh and Roundy [18] show the convexity of the objective function F (τ , γ ) under these conditions and present an efficient algorithm. In contrast, our capacity planning problem is not convex.…”
mentioning
confidence: 94%
“…The latter uses this policy to jointly optimize tool expansions along with nested floor and space expansions. Huh and Roundu [18] extend these ideas to a multi-product case under the Uniform Fill-Rate Production policy and identify a set of sufficient conditions for the capacity planning problem to be reduced to a nonlinear convex minimization program. This paper extends their model by introducing the layer of operations, the Lost Sales Cost Minimization allocation policy and tool retirement.…”
Capacity planning decisions affect a significant portion of future revenue. In equipment intensive industries, these decisions usually need to be made in the presence of both highly volatile demand and long capacity installation lead times. For a multiple product case, we present a continuous-time capacity planning model that addresses problems of realistic size and complexity found in current practice. Each product requires specific operations that can be performed by one or more tool groups. We consider a number of capacity allocation policies. We allow tool retirements in addition to purchases because the stochastic demand forecast for each product can be decreasing. We present a cluster-based heuristic algorithm that can incorporate both variance reduction techniques from the simulation literature and the principles of a generalized maximum flow algorithm from the network optimization literature.
pagesThis study addresses the problem of multi-period mix of product-lines under a product-family, which incorporates launching decisions of new products, capacity expansion decisions and product interdependencies. The problem is modelled as a two-stage stochastic program with recourse in which price, demand, production cost and cannibalisation effect of new products are treated as uncertain parameters.The solution approach employs the Sample Average Approximation based on Monte Carlo bounding technique and multi-cut version of L-shaped method to solve approximate problems efficiently, which is tested on different cases considering VSS and EVPI performance measures. The data collected through two experimental studies is analysed using ANOVA and Random Forest methodology in order to understand which problem parameters are significant on the performance measures and to generate some rule-based inferences reflecting the relationship between significant parameters and the performance of the proposed stochastic model.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.