Joseph F. Pekny

A simulation based optimization approach to supply chain management under demand uncertainty

Cost effective supply chain management under various market, logistics and production uncertainties is a critical issue for companies in the chemical process industry. Uncertainties in the supply chain usually increase the variance of profits (or costs) to the company, increasing the likelihood of decreased profit. Demand uncertainty, in particular, is an important factor to be considered in the supply chain design and operations. To hedge against demand uncertainty, safety stock levels are commonly introduced in supply chain operations as well as in supply chain design. Although there exists a large body of literature on estimating safety stock levels based on traditional inventory theory, this literature does not provide an effective methodology that can address the complexity of real CPI supply chains and that can impact the current practice in their design, planning and scheduling. In this paper, we propose the use of deterministic planning and scheduling models which incorporate safety stock levels as a means of accommodating demand uncertainties in routine operation. The problem of determining the safety stock level to use to meet a desired level of customer satisfaction is addressed using a simulation based optimization approach. An industrial-scale case problem is presented to demonstrate the utility of the proposed approach.

Enterprise-wide modeling & optimization—An overview of emerging research challenges and opportunities

The process systems engineering (PSE) as well as the operations research and management science (ORMS) literature has hitherto focused on disparate processes and functions within the enterprise. These themes have included upstream R&D pipeline management, planning and scheduling in batch and continuous manufacturing systems and more recently supply chain optimization under uncertainty. In reality, the modern process enterprise functions as a cohesive entity involving several degrees of cross-functional co-ordination across enterprise planning and process functions. The complex organizational structures underlying horizontally and vertically integrated process enterprises challenge our understanding of cross-functional co-ordination and its business impact. This article looks at the impact of enterprise-wide cross-functional coordination on enterprise performance, sustainability and growth prospects. Cross-functional coordination is defined as the integration of strategic and tactical decision-making processes involving the control of financial and inventory flows (both internal and external) as well as resource deployments. Initially, we demonstrate the existence of cross-functional decision-making dependencies using an enterprise network model. Subsequently, we discuss interactions between enterprise planning decisions involving project financing, debt-equity balancing, R&D portfolio selection, risk hedging with real derivative instruments, supply chain asset creation and marketing contracts which influence decision-making at the activity/process level. Several case studies are included to re-enforce the point that planning and process decisions need to be integrated.

Exact solution of large asymmetric traveling salesman problems.

The traveling salesman problem is one of a class of difficult problems in combinatorial optimization that is representative of a large number of important scientific and engineering problems. A survey is given of recent applications and methods for solving large problems. In addition, an algorithm for the exact solution of the asymmetric traveling salesman problem is presented along with computational results for several classes of problems. The results show that the algorithm performs remarkably well for some classes of problems, determining an optimal solution even for problems with large numbers of cities, yet for other classes, even small problems thwart determination of a provably optimal solution.

Risk and uncertainty in managing chemical manufacturing supply chains

Abstract A new metric is presented for evaluating supply chain design and planning projects in which there are significant elements of uncertainty and thus risk. The risk premium construct provides the basis for a rational balance between expected value of investment performance and variance. An effective polytope integration method for evaluation of expected values and variances of revenue is adopted which can account for the effects of demand uncertainties on revenue while recognizing the uncertainty in inventory over time. The combination of these elements with conventional deterministic mathematical programming models offers the promise of providing an effective approach to accommodating uncertainties and a rational basis for balancing risk. A small scale example is used to contrast the proposed approach with conventional stochastic programming-based methods. Another example shows the nature of the return and risk for a multiperiod production plan with stochastic effects on inventory. The computational complexities which are introduced by the risk premium construct are reviewed, and some directions for future research discussed.

A model predictive framework for planning and scheduling problems: a case study of consumer goods supply chain

Abstract Model Predictive Control is a well established technique for the control of processes and plants. We present a similar concept for planning and scheduling problems. There have mainly been two approaches to solve the planning and scheduling problems. The first approach is to model the planning and scheduling as one monolithic problem and solve it for the entire horizon. Needless to say, this approach requires an extensive computational effort and becomes impossible to solve in the case of large-scale scheduling problems. The other approach is to hierarchically-decompose the problem into a planning level problem and a scheduling level problem. This approach leads to tractable problems. Neither of these approaches provide the framework for incorporating uncertainties in the processing time of batches, or random equipment breakdowns, or demand uncertainties in the future. Furthermore, these approaches only provide ‘one snapshot’ of the planning problem and not a ‘walk through the timeline’. Model predictive planning and scheduling provides a framework for studying dynamics. Model predictive planning and scheduling requires a forecasting model and an optimization model. Both these models work in tandem in a simulation environment that incorporates uncertainty. The similarity with the model predictive approach which is widely used in process-control is that in each period, the forecasting model calculates the target inventory (controlled variable) in the future periods. These inventory levels ensure desired customer service level while minimizing average inventory. The scheduling model then tries to achieve these target inventory levels in the future periods by scheduling tasks (manipulated variables).

A simulation—optimization framework for addressing combinatorial and stochastic aspects of an R&D pipeline management problem

Abstract The R&D pipeline management problem has far-reaching economic implications for new-product-development driven industries, such as pharmaceutical, biotechnology, and agrochemical industries. Effective decision-making is required with respect to portfolio selection and project task scheduling in the face of significant uncertainty and an ever-constrained resource pool. In this paper, the here-and-now stochastic optimization problem inherent to the management of an R&D pipeline is described in its most general form. Subsequently, a computing architecture, Sim—Opt, is presented that combines mathematical programming and discrete event system simulation to assess the uncertainty and control the risk present in the pipeline. The concept of timelines, that studies multiple unique realizations of the controlled evolution of the discrete-event pipeline system, is introduced. Three different implementations of the decision-making module in Sim—Opt have been described and studied through an example case study.

Simulation-based optimization with surrogate models—Application to supply chain management

Simulation is widely used in the decision-making processes associated with supply chain management. In this paper, we present an extension of the simulation-based optimization framework which has been previously proposed for analyzing supply chains. The extension consists of the iterative construction of a surrogate model based on systematically accumulated simulation results to capture the causal relation between the key decision variables and supply chain performance. The decision variables can then be optimized using the surrogate model in place of individual simulation runs to economize on the overall computational effort. Several techniques are embedded in the framework to achieve the targeted objective: least square support vector machine (LSSVM), Bayesian evidence framework, and design and analysis of computer experiment (DACE). The extended framework is illustrated using two small examples and then applied to optimize the inventory levels in a three-stage supply chain. The results show that the framework identifies good solutions efficiently, can accommodate chance constraints and scales up well.

The curse of reality — why process scheduling optimization problems are difficult in practice

Abstract Two process scheduling examples from consumer goods industries are presented. These problems contain several features, which in practice tend to make problems difficult to solve and present barriers to regular use of scheduling technology. Several instances of the main challenge to obtaining a solution, the underlying process physics, are highlighted. Computational issues can be traced to operations being strongly interrelated by resource sharing, inventory constraints, manpower availability, and management policies. The question of what defines optimality arises when demands, process rates, yields, and batch cycle times are variable over time. Ultimately, it becomes necessary to tradeoff between optimality with respect to a given set of process parameter estimates and robustness to a set of realistic scenarios. Although the examples presented are relatively concise due to simplifying assumptions, implementation and representation of large problems are themselves challenges. If an enormous amount of resources are required to build and maintain models, even the highest quality methodology is in trouble.

Applying systems engineering principles in improving health care delivery

BackgroundIn a highly publicized joint report, the National Academy of Engineering and the Institute of Medicine recently recommended the systematic application of systems engineering approaches for reforming our health care delivery system. For this to happen, medical professionals and managers need to understand and appreciate the power that systems engineering concepts and tools can bring to redesigning and improving health care environments and practices.ObjectiveTo present and discuss fundamental concepts and tools of systems engineering and important parallels between systems engineering, health services, and implementation research as it pertains to the care of complex patients.DesignAn exploratory, qualitative review of systems engineering concepts and overview of ongoing applications of these concepts in the areas of hemodialysis, radiation therapy, and patient flow modeling.ResultsIn this paper, we describe systems engineering as the process of identifying the system of interest, choosing appropriate performance measures, selecting the best modeling tool, studying model properties and behavior under a variety of scenarios, and making design and operational decisions for implementation.ConclusionsWe discuss challenges and opportunities for bringing people with systems engineering skills into health care.

Perspectives on model based integration of process operations

Abstract The chemical process industry has increasingly been pursuing the use of computing technology to gather, organize, disseminate and exploit enterprise information and to closely coordinate the decisions made at the various levels of the process operational hierarchy so as to optimize overall corporate objectives. The thesis advanced in this paper is that mathematical programming models and solution methods offer the most effective tools for integration of the tactical and strategic levels of the operational hierarchy. To that end, existing strategies for implementing model-based integrated applications are reviewed. Four classes of examples of integration are presented: scheduling of multiproduct plants, large-scale model predictive control, integration of planning and scheduling across single and multiple plant sites and design of multipurpose batch plants under uncertainty. The methodology used to address these model-based integration instances successfully accomodates the key features of process operations: diverse time scale, multiple reference frameworks, spatial and organizational aggregation/disaggregation and uncertainty in the enterprise information. The applications further demonstrate that for the foreseeable future no single model or reference framework will be sufficient and efficient for treating all aspects of the process operations hierarchy.