This study is focused on two-sided assembly line balancing problems of type-I and type-II. This problem is encountered in production environments where a two-sided assembly line is used to produce physically large products. For type-I problems, there is a specified production target for a fixed time interval and the objective is to reach this production capacity with the minimum assembly line length used. On the other hand, type-II problem focuses on reaching the maximum production level using a fixed assembly line and workforce. Two different mathematical models for each problem type are developed to optimally solve the problems. Since the quality of the solutions by mathematical models decreases for large-sized problems due to time and memory limitations, two heuristic approaches are presented for solving large-sized type-I problem. The validity of all formulations is verified with the small-sized literature problems and the performances of the methods introduced are tested with large-sized literature problems.

Most classical unconstrained optimization methods require derivative information. Different methods have been proposed for problems where derivative information cannot be used. One class of these methods is heuristics including Evolutionary Algorithms (EAs). In this study, we propose EAs for unconstrained optimization under both deterministic and stochastic environments. We design a crossover operator that tries to lead the algorithm towards the global optimum even when the starting solutions are far from the optimal solution. We also adapt this algorithm to a stochastic environment where there exist only estimates for the function values. We design new parent selection schemes based on statistical grouping methods and a replacement scheme considering existing statistical information. We test the performance of our algorithms using functions from the literature and newly introduced functions and obtain promising results.

This study aims to investigate the outsourced design management implementations along the alliances that are established between firms and design consultancies in Turkey, which have been analysed in a limited sense in the past. Within this context / information on, &lsquo / why and how design is outsourced&rsquo / and &lsquo / how the design process is managed&rsquo / are derived through semi-structured interviews in conjunction with structured questionnaire conducted with managers from a sample selection of design and manufacturing firms with the aim of comparing the expectations and offerings of the parties that engaged the alliances in Turkey. The findings of the field study indicate that design management implementations in their current form of application exhibit several issues to discuss. Context of outsourced design service, corporate level strategic utilization of the design expertise, evaluation of final outputs, business initiation practices and design process management practices are important factors to consider in management of design alliances.

In manufacturing environments, machines are usually subject to down periods due to various reasons such as preventive maintenance activities, pre-accepted jobs and pre-known material shortages. Among these reasons, preventive maintenance, which is defined as the pre-planned maintenance activities to keep the machine in its operating state, has gained much more importance in recent years. In this thesis, we consider the single machine total flow time problem where the jobs are non-resumable and the machine is subject to preventive maintenance activities of known starting times and durations. We propose a number of optimality properties together with the upper and lower bounding procedures. Using these mechanisms, we build a branch and bound algorithm to find the optimal solution of the problem. Our extensive computational study on randomly generated test instances shows that our algorithm can solve large-sized problem instances with up to 80 jobs in reasonable times. We also study a two-alternative maintenance planning problem with minor and major maintenances. We give an optimizing algorithm to find the timing of the maintenances, when the job sequence is fixed.

In this study, the effect of inspection error on the product mix and quality projects selection in a manufacturing environment where rework and inspection errors exist is examined. It is assumed that the products (items) for which rework is necessary are reprocessed at a separate work center and 100% inspection is performed for the products both after rework and processing operations. Markov chain approach is used to compute yield and rework rates. In addition, nominal-the-best type of a quality loss function is used in computing quality loss due to products shipped to the customers. A linear programming (LP) model is developed to support the product mix and quality improvement project selection decisions. The use of LP model is demonstrated on an example problem. The results obtained under different experimental conditions are compared with solutions of a naive QI project selection method, improving the least capable process. The analysis shows that developed LP model is relatively better than process capability approach. Besides, according to the results obtained under different experimental conditions, the factors that have significant effect on throughput and QI project selection are being determined.

In this study, non-permutation flowshops with missing operations are considered. The primary performance criterion is the total cycle time (i.e. makespan) and the secondary criterion is the total flowtime. In order to obtain the schedule with minimum makespan and minimum total flowtime, non-permutation schedules are being generated instead of permutation ones by permitting multiple jobs bypassing stages where misisng operations occur. A heuristic algorithm has been developed in order to generate non-permutations sequences through those stages. The heuristic algorithm has been compared with the existing heuristic methods in the literature, the ones generating permutation vs. the ones generating non-permutation schedules. Computational analysis is conducted to investigate the effects of certain parameter values such as the number of machines, the number of jobs and the percentage of missing operations. The results demonstrate slight improvement in the makespan as well as a significant improvement in total flowtime of schedules generated by the new heuristic procedure compared to leading non-permutation and permutation schedule generating heuristics, where the percentage of improvement gets higher with larger percentage of missing operations.

Manufacturing companies are searching for new techniques day by day to launch qualified products with a competitive price to the market. Companies believe in the importance of teams which have members coming from different disciplines who use their knowledge, experiences, and creativity for achieving the goals of their teams. The purpose of this study is to explore the effectiveness of teamwork for successful product development process by focusing on the relationship between the team members &ndash / especially engineers&rsquo / and industrial designers&rsquo / relation. After a broad literature survey, a descriptive-survey study that aims demonstrating the relationship between engineers and industrial designers in manufacturing companies in Turkey is held in order to investigate the effectiveness of teamwork in product development process. A twelve item questionnaire which has four open-ended questions and eight Likert-scaled statements has been prepared. As the result of this study, a positive relationship is found between the performance of product development team and the effects of cooperation, social and professional communication, having clear and common goals, sharing knowledge and experiences, leadership, coordination and cooperative problem solving and decision making processes to the relationship between engineers and industrial designers.

In some production environments the defect rates are considerably low such that measurement of fraction of nonconforming items reaches parts per million level. In such environments, monitoring the number of conforming items between consecutive nonconforming items, namely the time between events (TBE) is often suggested. However, in the design of control charts for TBE monitoring a common practice is the assumptions of known process parameters. Nevertheless, in many applications the true values of the process parameters are not known. Their estimates should be determined from a sample obtained from the process at a time when it is expected to operate in a state of statistical control. Additional variability introduced through sampling may significantly effect the performance of a control chart. In this study, the effect of parameter estimation on the performance of Time Between Events Exponentially Weighted Moving Average (TBE EWMA) schemes is examined. Conditional performance is evaluated to show the effect of estimation. Marginal performance is analyzed in order to make recommendations on sample size requirements. Markov chain approach is used for evaluating the results.

In this thesis, a real world application of the constant work in process (Conwip) system in an electronics company is realized. The aim of the application is to reduce the work in process (WIP) inventory while maintaining the same throughput level. A model is developed to determine the constant work in process level of the Conwip system for the production lines in this company. The approximated mean value analysis approach is used for the solution. Real system data are collected before and after the Conwip application. Hypothesis tests are used to compare the WIP and the throughput levels of the Conwip system with the existing push control system for a pilot production line. Results of the hypothesis tests show that the Conwip production control system can signinifcantly reduce the WIP while maintaining the same throughput rate.

We consider a manufacturing system with stochastic demand and predictable production yield. The manufacturer has predetermined prices and limited production capacity in each period. The producer also has the option to save some inventory for future periods even if there is demand in the current period. The demand that is not met is lost or may be backordered for only one period. Our objective is to maximize the expected profit by choosing optimal production, save and backorder quantities in each period. We formulate this problem as a Markov Decision Process where the state of the system is represented by the net inventory and the efficiency parameter. We show that a modified (Y, S, B) policy is optimal in each period. At the end, we have some computational analysis to examine the effects of yield on the optimal decisions.