Manufacturing systems are becoming more and more flexible to confront uncertainty in demands of multiple product types. In this paper, we study the properties of flexible manufacturing systems from the perspective of demand shortfall and service level as functions of setup time, nonhomogeneous capacities, and demand distributions. Closed-form bounds as well as an accurate estimate of the service level are obtained. Sensitivity analyses of service level with respect to reductions in setup time, downtime, and cycle time are carried out, and the impact of new product launch and product correlation on system flexibility is investigated.

This paper is devoted to modeling, analysis, and improvement of flexible manufacturing systems with asynchronous exponential machines and finite nondedicated buffers. In such systems, each machine may process multiple types of products with different speeds, and buffers are shared for all products. Efficient algorithms to evaluate the system performance are developed. Formulas are derived to calculate line throughput in one- and two-machine lines, and a convergent recursive algorithm is introduced for longer lines. The numerical results show that the method leads to a high accuracy in performance evaluation. Using such a model, bottleneck analysis has been carried out to identify the machine or product whose improvement will lead to the largest improvement in system throughput. Indicators based on the collected data to identify bottleneck machine and product are derived without complete calculation of the partial derivatives of system performance. Such efficient algorithms provide a quantitative tool for analysis and improvement of flexible manufacturing systems.

Improving quality in large volume battery manufacturing systems for hybrid and electric vehicles is of significant importance. In this paper, we present a flow model to analyze and improve product quality in electrical vehicle battery assembly lines with 100% inspections and repairs for defective parts. Specifically, a battery assembly line consisting of multiple inspection stations is considered. After each inspection, defective parts will be repaired and sent back to the line. A quality flow model is introduced to analyze quality propagations along the battery production line. Analytical expressions of final product quality are derived and structural properties, such as monotonicity and sensitivities, are investigated. A bottleneck identification and mitigation method is introduced to improve quality performance. Finally, a case study is presented to illustrate the applicability of the method.

This paper explores energy efficient building design for manufacturing plants. Many efforts have been directed into the field of building design optimization concerning building energy performance, but most of the studies focus on residential buildings or public buildings. Very limited research results studying plants buildings have been reported. However, plants buildings have certain unique features that make the design problem more challenging. Furthermore, the approaches presented in the current publications could not guarantee the performance of their designs if the computation capacity is limited. This paper attempts to address these two issues. First, an EnergyPlus-integrated overall energy consumption estimation framework is developed for a class of manufacturing plants, where the environmental conditions would not affect the energy consumption of the production processes. Based on that, the building design problem for this type of manufacturing plants is formulated as a stochastic programming problem concerning uncertainties arising from the future weather conditions and energy prices, where seasonal production scheduling optimizing is incorporated when estimating the performance of building designs. Second, Ordinal Optimization (OO) method is introduced to solve the problem so as to quantitatively guarantee a high probability of finding satisfactory designs while reducing the computation burden. A numerical example is provided, showing our solution method performs effectively in finding a satisfactory design.

In this paper, we present an analytical method to evaluate the quality performance of flexible manufacturing systems with batch operations. By using a Markov chain model, a closed formula to quantify the probability of producing a good part is derived and nonmonotonic properties in quality are investigated.

In systems of general assembly line with material handling, line-side buffers need to be carefully assigned to a limited number of material delivers (drivers) for part delivery to avoid production stoppage due to material shortage. Such a problem is referred to as line-side buffer assignment problem (LBAP). In this paper, we focus on fixed zoning version of LBAP. We formulate the problem, prove its NP-hardness, and propose an algorithm based on two structural characteristics of the LBAP problem-one being the analogousness between our problem and the parallel machine scheduling (PMS) problem and the other being the monotonicity of the system throughput in the course of assigning line-side buffers to drivers. The developed algorithm globally converges with probability one when there exist feasible assignments. The algorithm is tested on a real system, and the results show that it is effective for solving the LBAP problem.

Performance evaluation of complex manufacturing systems is challenging due to many factors such as system complexity, parameter uncertainties, problem size, just to name a few. In many cases when a system is too complex to model using mathematical formulas, simulation is used as an effective alternative to conduct system analysis. A manufacturing system is a good example of such cases where both system performance and system complexity are greatly impacted by material handling (MH) strategy, management, and operational control. In this paper, we study vehicle general assembly (GA) system with MH, and focus on developing an efficient simulation method for modeling and analysis where traditional simulation methods may suffer from computation intensity. Making use of the partial system decomposability, we introduce an aggregated event-scheduling simulation method with two-level framework. A dividing mechanism with boundary conditions is employed in top-level simulation to divide the global event list into small sizes. A timing-focuses strategy based on max-plus algebra is applied in bottom-level local simulation to further reduce local event lists. With this new method it is possible to mimic real production systems fast and accurately within a reasonable computational time frame. The effectiveness and efficiency of the new simulation method are validated through experimental results.

Most of the studies in flexible manufacturing systems address the issues of flexibility, productivity, cost, etc. The impact of flexible lines on product quality is less studied. This paper presents a quantitative model based on Markov chain analysis to evaluate quality performance of a flexible machining system. A case study of flexible fixture is provided to illustrate the applicability of the method.

This paper presents an iterative approach to model and analyze one type of flexible manufacturing systems with common lines and dedicated branches to process different products. The mathematical procedures, associated with the justification of convergence and accuracy, are provided. A case study is presented to illustrate the applicability of the method.