A probability based model for evaluating delivery performance.

Author:Guiffrida, Alfred L.

    Organizations are under constant pressure to shorten product lifecycles, reduce costs and improve customer service in order to gain and maintain competitive advantage. Supply chain management has evolved as a core component of an organization's overall competitive strategy for attaining competitive advantage. Supply chain management serves as the foundation for integrating and effectively managing value-adding activities such as raw materials acquisition, production processing and physical distribution. A supply chain management based operating philosophy requires an organization to adoption a formal performance measurement system that has a diversified set of performance measures that can assist managers in meeting day-to-day as well as long term performance goals. Performance measures that accurately reflect supply chain operations are required to support continuous improvement activities within the organization. The importance and integration of performance measurement systems within supply chains has been addressed by several researchers (see for example Martin and Patterson, 2009; Gunasekaran and Kobu, 2007; and Tan et al. 2002).

    In this research we concentrate on one aspect of overall supply chain performance, delivery timeliness to the final customer. Recent research has identified delivery performance as a key management concern among supply chain managers (see for example Lockamy and McCormack, 2004; Min and Zhou, 2002; and Vachon and Klassen, 2002). Conceptual frameworks for defining delivery performance in supply chain management are found in Gunasekaran et al. (2001) and Fawcett et al. (1997). Within these frameworks, delivery performance is classified as a strategic level supply chain performance measure.

    1.1 Delivery Performance Models for Multi-stage Supply Chains

    Models for evaluating delivery performance to the final customer within multi-stage supply chains have been proposed by several researchers. These models can be classified according to two distinct mathematical modeling approaches. The first class of models employs capability indices to model delivery performance. Garg et al. (2006) utilize a six-sigma statistical design tolerance methodology and create a "delivery capability index" that is similar to structure to the Cpk process capability index that is used manufacturing. The index is used to optimally distribute the pool of activity variance that results when manufacturing a product in a multi-stage supply chain so as to satisfy customer delivery expectations with respect to a delivery window which designates early, on-time and late deliveries. Wang and Du (2007) develop a similar six-sigma driven delivery capacity indexing method. They define a total cost model for evaluating delivery performance subject to a customer defined delivery window and demonstrate how the model can be used by supply chain managers to make supplier selection decisions.

    A second class of supply chain delivery performance models are cost-based decision models that advocate improving delivery performance (subject to a customer defined delivery window) by reducing the variance of the delivery distribution. Guiffrida et al. (2008) and Guiffrida and Jaber (2008) present budget constrained nonlinear optimization models which capture the expected costs for early and late delivery. Bounds for justifying financial investment for improving on-time delivery performance are established and the impact of failing to invest in improving supply chain delivery performance is financial quantified as the opportunity cost of "managerial neglect" (Guiffrida and Nagi, 2006).

    The two classes of aforementioned models are elegant in their application of statistical theory and mathematical optimization to the task of evaluating supply chain delivery performance. These models require that the end user have a somewhat sophisticated background in statistics and decision theory. In practice, a delivery performance model that improves delivery performance by reducing the probabilities of early and late deliveries is easier for a practitioner to understand and can be more easily integrated into current supplier evaluation programs in industry which often track the frequency of untimely deliveries. In this paper we present a supply chain delivery performance model wherein delivery performance is measured in the easily understood metric of the probability of early, on-time and late delivery. This paper is organized as follows. In Section 2 we introduce the mathematical form of a model for evaluating supply chain delivery performance subject to a customer defined delivery window. In Section 3, three approximate solution methodologies for evaluating delivery performance are formulated. A set of numerical comparisons are presented and...

To continue reading