Emre A. Veral

Computer simulation of due-date setting in multi-machine job shops

The due-date setting and operations scheduling problem is addressed in an unbalanced, multi-machine random job shop. The focus of the study is to demonstrate the feasibility of setting reliable static due-dates through operation flow time analysis. Response-surface mapping methodology via regression analysis is employed to model operation flow time characteristics, which are shown to be non-linear and dispatching rule dependent. Discussion points out the advantages and practicality of using static job information as opposed to dynamic shop information in setting due-dates. Simulation results and statistical analyses show the viability of setting due-dates that are tight, reliable, and consistent, using this methodology. A unique characteristic of the proposed method is that it shows simultaneous reductions in variability of manufacturing lead times, tardiness, proportion of tardy jobs, and maximum tardiness without resorting to looser due-dates.

Assessing appointment systems’ operational performance with policy targets

We propose a paradigm shift in how the performance of outpatient clinic appointment schedules is evaluated in practice and academia. Our research addresses the traditional dilemma between patients’ wait times and providers’ idle time and overtime, but with operational performance metrics that assess their respective probabilities of exceeding established thresholds, instead of optimizing a presumed cost function. Using a stochastic model, we introduce a new way of analyzing appointment schedules that is absent from the literature but appealing to practitioners. We take into account the variable nature of patient consultation times, known differences in the duration of diverse consults, and patients’ propensity to miss their appointments. Analysis shows that traditional scheduling systems have serious shortcomings in terms of providing consistent service levels, and we conclude that the managerial decision space so far investigated in the appointment scheduling literature is not adequate for exercising operational control over appointment system performance.

Designing appointment system templates with operational performance targets

We investigate characteristics of outpatient appointment templates that provide fair and controlled patient waiting experiences. Using an inverse-simulation application, we establish successive patients’ designated arrival times based on the previous patient’s finish time distribution and a universal targeted wait time set by the provider. This approach results in a template where each patient’s probability of waiting longer than a threshold duration is uniform across all patients. Through investigation of various no-show probabilities and service time distributions, we show that the template designs that achieve wait time uniformity across different patient and service environments have generalizable characteristics. Based on these characteristics, we proceed to design practitioner-oriented heuristics that create rounded interval times for implementation which yield robust performance outcomes. Results suggest that such schedules consist of appointment intervals that differ from average service times, moderated by patient show-rates and service time characteristics. In addition to introducing methods that allow patient “service level agreements,” our conclusions bring into question the advisability of double-booking, multiple-block scheduling, and yield management practices in appointment scheduling, and provide support for interval adjustment approaches that respond to patient service and no-show characteristics.

A two-phased approach to setting due-dates in single machine job shops

This paper explores the characteristics of job flow time distributions resulting from the use of different priority rules via computer simulation and statistical analysis. Statistical analysis of results from phase I experimentation modeling traditional due-date setting practices are used to develop job flow time prediction models. In phase II simulation experiments, the regression models developed in phase I are used to establish due-date allowances to jobs based on their processing times. Forms of the prediction equations reveal the differences in resultant job flow distributions, and explain the underlying nature of interactions between due-date setting methods and shop performance characteristics of dispatching rules as reported in previous studies. Application of the prediction equations in phase II experimentation show the viability of improving shop floor performance without resorting to looser due-dates.

Using production and transfer batches in flowshops under MRP control

Abstract This study investigates the impact of using material flow control and order release throttling mechanisms on shop floor effectiveness in MRP based order release environments. The research methodology involves simulation experiments conducted using a flow shop model of a manufacturing and assembly setting. Specifically addressed issue is the decoupling of shop floor control decisions from MRP release schedules. This decoupling effect is implemented through the use of smaller production batch sizes, staggering the release of the production quantities to initial work stations, and using transfer batches. In a 14 parts/9 processes simulation environment, the effects of shop floor control decisions are evaluated based on average hours of work in process inventory and average number of tardy units for the simulated planning horizon. Experimental results suggest that increasing material flow frequency through smaller process and transfer batch sizes have significant consequences on inventory and service levels. By separating work in process inventories into two components, it is shown that reducing batch sizes may not affect inventory levels of asynchronized assembly units (i.e. component delay). The results also indicate the presence of a possible threshold effect for throttling and staggering work order releases where service levels are compromised and marginal gain on inventory reduction diminishes.

The Applications of Revenue Management and Pricing

Revenue Management (RM) and pricing may be described as the art of selling products to the right customers at the right prices. The concept is based on the assumption that different customers are willing to pay different prices for the same product and that by differentiating the price according to customer characteristics, overall revenue can be maximized. While the term “Revenue Management” often refers to the problem of defining the amount of products to be offered at one price, the term “pricing” usually refers to the problem of defining optimal prices. Historically (Bobb and Veral, 2008), Revenue Management started as an operations management function, focusing only on capacity allocation given exogenous demand estimates (Gallego and Van Ryzin, 1997). In the 1960s, American Airlines started to use Operations Research models for Revenue Management decisions. Littlewood (1972) presented the revenue maximization model through booking limits and inventory control systems. The 1980s saw Revenue Management become a robust and workable system for solving the problems of fixed capacity, time-varied demand, segmentation, perishable inventory and high fixed costs, bringing a practical solution. One significant milestone in Revenue Management was Peter Belobaba’s PhD thesis work, Air Travel Demand and Airline Seat Inventory Management (1987) which was a significant contribution to management of complexity, capacity allocation and real-time inventory solutions.

Mattel, Inc., Global Manufacturing Principles (GMP): A Life-Cycle Analysis of a Company’s Voluntary Code of Conduct

In November 1997, Mattel announced the creation of a global code of conduct for its production facilities and contract manufacturers. Called the Global Manufacturing Principles (GMP), the code covered such issues as wages and hours, child labor, forced labor, discrimination, freedom of association, legal and ethical business practices, product safety and product quality, protection of the environment, and respect for local cultures, values, and traditions.

MRP strategies: Effects on cash flow

Abstract As material requirements planning systems do not consider capacity limitations in their inherent logic, considerable effort must be spent to achieve feasible production plans. This simulation study examines the cash flow consequences of nine production planning strategies consisting of three master production scheduling methods and three component/part scheduling methods. Experimentation examines different operational cost ratios. Evaluation criteria used are characterized by total cash flow and periodic variablity of cash flow to support operating expenses for each planning strategy. Simulation results indicate that sophisticated methods for planning do not necessarily cause significant savings. A second finding is that a smoother master production schedule tends to result in more stable cash outflow patterns. In comparison, component scheduling efforts are found to have relatively insignificant consequences when compared to the effects of master production schedules.