Search

Completion time requirements are often imposed on a classification task. In practice, the desired completion time for classifying a subject may depend on its label (target) value. For example, a [...]

An external contractor receives a payment from a manufacturer for periodically performing preventive maintenance and for performing minimal repairs whenever process failures occur. Suppose a new technology will be available in the future, but the timing is uncertain. We propose specifying one or multiple time points in a maintenance contract, at which the manufacturer can adopt (switch to) the new technology, if it becomes available, and change the preventive maintenance schedule for the remaining time in the contract period. A model is developed to determine the value of using these (switch) points in a maintenance contract to the manufacturer. The numerical results show that, with a higher arrival rate of the new technology, the value becomes greater, and thus, the optimal switch point should be set earlier. Furthermore, if the new technology reduces maintenance costs, using switch points adds more value to a cost-plus-margin method than either the fixed payment or cost-plus-fixed-fee methods. On the other hand, if adopting the new technology results in higher maintenance costs, the value using switch points is higher for the fixed payment method. Index Terms--Contract valuation, maintenance contract, technology adoption.

Consider a manufacturer who has a process with an increasing failure rate over time. In order to improve the process performance, the following two types of maintenance activity are outsourced to an external contractor: (i) preventive maintenance is performed periodically to improve the reliability of the process when the process is functional; and (ii) corrective maintenance is used to restore the process to a specified condition when it fails. We consider the use of incentive contracts to induce the contractor to select the maintenance policy that optimizes the total profit of the manufacturer and contractor. It is demonstrated that an incentive contract based on a combination of a target uptime level and a bonus always leads to the desired win-win coordination, and provides flexibility in allocating the extra profit generated from coordination and, importantly, an incentive to the contractor to improve the efficiency of the maintenance operations. The incentive contract can also be used to select the most economically efficient contractor from multiple contractors with different maintenance capabilities., 1. Introduction Effective channel coordination is essential if an organization is to succeed in today's highly competitive business environment. In the last two decades, channel coordination issues in supply chains [...]

Consider a manufacturer with a manufacturing system that consists of multiple processes. The manufacturer's revenue is determined by the minimum of the uptimes among the processes. The maintenance functions of the processes are outsourced to independent contractors so that each contractor is responsible only for one process. A performance-based incentive contract is offered to each contractor, consisting of an uptime target level and a bonus rate for exceeding the uptime target. Under the incentive contract, a contractor receives a bonus only when the achieved uptime exceeds the target level specified in the contract. We develop a model for jointly determining the uptime target levels and bonus rates for the contractors which maximize system profit. We also demonstrate the financial benefits of coordination and added flexibility in allocating the additional profit to the contractors. In addition we show the impact of the variation in individual process maintenance times and costs on channel coordination and profits., 1. Introduction Traditionally, maintenance has been a core in-house activity. However, it is now being viewed by many organizations as a non-core business function, and outsourcing of the maintenance function [...]

A study on the process control policies with the help of a process control model is presented which helps in studying the policies when the value of the product. The model is based on internal and external failure costs and is developed under a time-varying cost structure.

In the so-called two-attribute warranty policy, two types of warranty criteria, such as the age and mileage of an automobile, are employed simultaneously to determine the eligibility of a warranty claim. We propose in the paper several decision models that estimate the expected total cost incurred under various types of two-attribute warranty policies. We also perform a sensitivity analysis to study the effects of several model parameters, such as the discount rate, the product usage rate, and the warranty terms, on the total warranty cost. Index Terms - Product failure process, two-dimensional warranty, warranty cost.

A two-echelon model was developed to identify the process mean in a case in which quantity discounts are present in the raw material acquisition cost. The model developed tries to integrate concepts related with production run size and raw material procurement policy into a single-product production process mean problem. A solution procedure was generated by the model for a joint identification of the process mean, the production setup and the raw material ordering policies..

A linear time algorithm was developed for analyzing the flow time of each arc that is developed within the time-window constraint and the time-schedule constraint affecting project management. The linear algorithm was developed by incorporating the two types of constraint in the traditional activity network. Furthermore, the solution method was characterized as a modification of the forward pass, backward pass algorithm used for traditional activity networks.

A model for a single production process involving a constant supply of raw material offered by a reliable source was developed and analyzed to develop optimal solutions for quality control. The two-echelon model which exhibited three cases involving quantity discounts in the raw material costs incorporated issues associated with production run size and raw material procurement policy into the classical process mean problem. Furthermore, the model provided a useful framework for developing efficient procedures for a joint determination of production setup.

The use of on-line sensors in maintaining a randomly deteriorating machine is considered. Specifically, a steady state threshold policy and a Bayes-based heuristic for determining when to re-set based on sensor measurements are developed and compared. Two assumptions were considered, namely, that the deterioration follows a time-invariant Markov process and that the machine condition may be modeled by a finite number of states.

The problem of finding the optimal warranty price for a free-replacement, fixed-period warranty policy for a certain warranty period is investigated. Specifically, a decision model for maximizing the manufacturer's profit equivalent is developed based on the producer's and customers' risk preferences, their views on product failure rate, the customer's loyalty to the producer in repairs, and the customers' repair price.

A two-stage inspection process was developed to determine whether manufactured products meet product specifications. If they do, then such products can be sold to markets which offer high-value. If the products do not meet specifications, then they are sold for scrap at lower value. Tests of the two-stage process revealed that it is capable of setting optimal screening specification limits for both known and unknown parameters.

An asymptotic expansion, used to calculate Bayes credibility intervals, was obtained for the posterior Cdf of the reliability of series systems containing very reliable components. Bayes credibility intervals were calculated when the infinite series expansion was truncated. The asymptotic expansion was used to determine the magnitude of error when the infinite series expansion was eliminated.

Two cost models are developed to estimate the optimal burn-in time of products to identify early failures before shipment at minimum per-item expected cost. A Weibull-exponential distribution was used as a product life distribution model. Results show that, for both repairable and non-repairable products, burn-in time should not exceed that of the change-point, which can be estimated by the maximum-likelihood method.

In this paper, we present different inspection methods to determine the acceptability of produced items. In the first method, items are classified into two grades. An item belongs to a grade when it meets the requirements of that grade. In the second method, a joint screening rule based on an aggregation of the characteristics is used. The two methods are then compared. [ABSTRACT FROM AUTHOR]