Assessing process capability based on Bayesian approach with subsamples

Process capability indices (PCIs) have been widely used to measure the actual process information with respect to the manufacturing specifications, and become the common language for process quality between the customer and the supplier. Most of existing research works for capability testing are based on the traditional frequentist point of view and statistical properties of the estimated PCIs are derived based on the assumption of one single sample. In this paper, we consider the problem of estimating and testing process capability using Bayesian approach based on subsamples collected over time from an in-control process. The posterior probability and the credible interval for the most popular index _Cpk$C_{\mathit{pk}}$ under a non-informative prior are derived. The manufacturers can use the presented approach to perform capability testing and determine whether their processes are capable of reproducing product items satisfying customers’ stringent quality requirements when a daily-based or weekly-based production control plan is implemented for monitoring process stability.     

Variable sampling inspection for resubmitted lots based on process capability index Cpk for normally distributed items

In this article, a variables sampling inspection plan for resubmitted lot based on the process capability index C_pk is developed for normally distributed items with unknown mean and variance. The three plan parameters, sample size n, critical acceptance value k_a and the number of resubmissions m, are determined simultaneously by minimizing average sample number (ASN) with two conditions specified by the producer and the consumer. The advantages of the proposed resubmitted sampling plan over the variables single sampling plan are also discussed. Tables of plan parameters for selected values of acceptable quality level (AQL), limiting quality level (LQL), producer’s α-risk and consumer’s β-risk are provided and discussed with the help of an application example.     

Decision-making in testing process performance with fuzzy data

Over the years, numerous process capability indices (PCIs) have been proposed to the manufacturing industry to provide numerical measures of process performance. Most research efforts have focused on developing and investigating PCIs that assess process capability by precise measurements of output quality. However, real observations of continuous quantities are not precise numbers; in practice, they are more or less imprecise. Since observations of continuous random variables are imprecise the values of related test statistics become imprecise. Therefore, decision rules for statistical tests have to be adapted to this situation. This article presents a set of confidence intervals that produces triangular fuzzy numbers for the estimation of C_pk index using Buckley’s approach with some modification. Additionally, a three-decision testing rule and step-by-step procedure are developed to assess process performance based on fuzzy critical values and fuzzy p-values. This concept is also illustrated with an example for testing process performance.     

A Bayesian approach for assessing process precision based on multiple samples

Using process capability indices to quantify manufacturing process precision (consistency) and performance, is an essential part of implementing any quality improvement program. Most research works for testing the capability indices have focused on using the traditional distribution frequency approaches. Cheng and Spiring [IIE Trans. 21 (1) 97] proposed a Bayesian procedure for assessing process capability index C_p based on one single sample. In practice, manufacturing information regarding product quality characteristic is often derived from multiple samples, particularly, when a routine-based quality control plan is implemented for monitoring process stability. In this paper, we consider estimating and testing C_p with multiple samples using Bayesian approach, and propose accordingly a Bayesian procedure for capability testing. The posterior probability, p, for which the process under investigation is capable, is derived. The credible interval, a Bayesian analogue of the classical lower confidence interval, is obtained. The results obtained in this paper, are generalizations of those obtained in Cheng and Spiring [IIE Trans. 21 (1), 97]. Practitioners can use the proposed procedure to Cheng and Spiring determine whether their manufacturing processes are capable of reproducing products satisfying the preset precision requirement.     

Production yield measure for multiple characteristics processes based on {S_{pk}^T} under multiple samples

Process yield is an important criterion used in the manufacturing industry for measuring process performance. Methods for measuring yield for processes with single characteristic have been investigated extensively. However, methods for measuring yield for processes with multiple characteristics have been comparatively neglected. Chen et al. (Qual Reliab Eng Int 19:101–110, 2003) proposed a measurement formula called S_pk^T{S_{pk}^T } , which provides an exact measure of the overall process yield, for processes with multiple characteristics. In this paper, we considered the natural estimator of S_pk^T{S_{pk}^T } under multiple samples, and derived the asymptotic distribution for the estimator. In addition, a comparison between the SB (standard bootstrap) and the proposed method based on the lower confidence bound is executed. Generally, the result indicates that the proposed approach is more reliable than the standard bootstrap method.     

Cpk index estimation using fuzzy numbers

Process capacity indices (PCIs) were developed and have been successfully used by companies to compete in and dominate the high-profit markets by improving the quality and the productivity since the past two decades. There is an essential assumption, in the conventional application, wherein the output process measurements are precise and distributed as normal random variables. Since the assumption of normal distribution is untenable, errors can occur if the C_pk index is computed using non-normal data. In the present study, we address the situation that the output of data from measurement of the quality of a product is insufficiently precise or scarce. This is possible when the quality measurement refers to the decision-maker’s subjective determination. In such a situation, the linguistic variable that is easier to capture the decision-maker’s subjective perception is applied to construct the PCI C_pk. The present approach can mitigate the effect when the normal assumption is inappropriate and extends the application of C_pk index.     

Lower Bounds for Scheduling on Identical Parallel Machines with Heads and Tails

In this paper, we investigate new lower bounds for the P|r _j ,q _j |C _max scheduling problem. A new bin packing based lower bound, as well as several new lifting procedures are derived for this strongly NP -hard problem. Extensive numerical experiments show that the proposed lower bounds consistently outperform the best existing ones.     

Optimal designs for weighted approximation and integration of stochastic processes on [0,∞)

We study minimal errors and optimal designs for weighted L₂-approximation and weighted integration of Gaussian stochastic processes X defined on the half-line [0,∞). Under some regularity conditions, we obtain sharp bounds for the minimal errors for approximation and upper bounds for the minimal errors for integration. The upper bounds are proven constructively for approximation and non-constructively for integration. For integration of the r-fold integrated Brownian motion, the upper bound is proven constructively and we have a matching lower bound.     

Code constructions and existence bounds for relative generalized Hamming weight

The relative generalized Hamming weight (RGHW) of a linear code C and a subcode C ¹ is an extension of generalized Hamming weight. The concept was firstly used to protect messages from an adversary in the wiretap channel of type II with illegitimate parties. It was also applied to the wiretap network II for secrecy control of network coding and to trellis-based decoding algorithms for complexity estimation. For RGHW, bounds and code constructions are two related issues. Upper bounds on RGHW show the possible optimality for the applications, and code constructions meeting upper bounds are for designing optimal schemes. In this article, we show indirect and direct code constructions for known upper bounds on RGHW. When upper bounds are not tight or constructions are hard to find, we provide two asymptotically equivalent existence bounds about good code pairs for designing suboptimal schemes. Particularly, most code pairs (C, C ¹) are good when the length n of C is sufficiently large, the dimension k of C is proportional to n and other parameters are fixed. Moreover, the first existence bound yields an implicit lower bound on RGHW, and the asymptotic form of this existence bound generalizes the usual asymptotic Gilbert–Varshamov bound.     

Bounds for solid angles of lattices of rank three

We find sharp absolute constants C₁ and C₂ with the following property: every well-rounded lattice of rank 3 in a Euclidean space has a minimal basis so that the solid angle spanned by these basis vectors lies in the interval [C₁,C₂]. In fact, we show that these absolute bounds hold for a larger class of lattices than just well-rounded, and the upper bound holds for all. We state a technical condition on the lattice that may prevent it from satisfying the absolute lower bound on the solid angle, in which case we derive a lower bound in terms of the ratios of successive minima of the lattice. We use this result to show that among all spherical triangles on the unit sphere in RN with vertices on the minimal vectors of a lattice, the smallest possible area is achieved by a configuration of minimal vectors of the (normalized) face centered cubic lattice in R³. Such spherical configurations come up in connection with the kissing number problem.     

Improved upper and lower bounds on the feedback vertex numbers of grids and butterflies

We improve upon the best known upper and lower bounds on the sizes of minimal feedback vertex sets in butterflies. Also, we construct new feedback vertex sets in grids so that for a large number of pairs (n,m), the size of our feedback vertex set in the grid M_n,m matches the best known lower bound, and for all other pairs it differs from this lower bound by at most 2.     

Fuzzy inference to assess manufacturing process capability with imprecise data

Process capability indices provide numerical measures on whether a process conforms to the defined manufacturing capability prerequisite. These have been successfully applied by companies to compete with and to lead high-profit markets by evaluating the quality and productivity performance. The loss-based process capability index C_pm, sometimes called the Taguchi index, was proposed to measure process capability, wherein the output process measurements are precise. In the present study, we develop a realistic approach that allows the consideration of imprecise output data resulting from the measurements of the products quality. A general method combining the vector of fuzzy numbers to produce the membership function of fuzzy estimator of Taguchi index is introduced for further testing process capability. With the sampling distribution for the precise estimation of C_pm, two useful fuzzy inference criteria, the critical value and the fuzzy P-value, are proposed to assess the manufacturing process capability based on C_pm. The presented methodology takes into the consideration of a certain degree of imprecision on the sample data and leads to the three-decision rule with the four quadrants decision-making plot. The fuzzy inference procedure presented to assess process capability is a natural generalization of the traditional test, when the data are precise the proposed test is reduced to a classical test with a binary decision.     

A variables sampling plan based on Cpmk for product acceptance determination

Process capability indices are useful management tools, particularly in the manufacturing industry, which provide common quantitative measures on manufacturing capability and production quality. Most supplier certification manuals include a discussion of process capability analysis and describe the recommended procedure for computing a process capability index. Acceptance sampling plans have been one of the most practical tools used in classical quality control applications. It provides both vendors and buyers to reserve their own rights by compromising on a rule to judge a batch of products. Both sides may set their own safeguard line to protect their benefits. Two kinds of risks are balanced using a well-designed sampling plan. In this paper, we introduce a new variables sampling plan based on process capability index C_pmk to deal with product sentencing (acceptance determination). The proposed new sampling plan is developed based on the exact sampling distribution hence the decisions made are more accurate and reliable. For practical purpose, tables for the required sample sizes and the corresponding critical acceptance values for various producer’s risk, the consumer’s risk and the capability requirements acceptance quality level (AQL), and the lot tolerance percent defective (LTPD) are provided. A case study is also presented to illustrate how the proposed procedure can be constructed and applied to the real applications.     

Resolution Complexity of Random Constraint Satisfaction Problems: Another Half of the Story

Let C_n,cn^2,k,t be a random constraint satisfaction problem(CSP) of n binary variables, where c > 0 is a fixed constant and the cn constraints are selected uniformly and independently from all the possible k-ary constraints each of which contains exactly t tuples of the values as its restrictions. We establish upper bounds for the tightness threshold for C_n,cn^2,k,t to have an exponential resolution complexity. The upper bounds partly answers the open problems regarding the CSP resolution complexity with the tightness between the existing upper and lower bound [1].     

Bounds for the first positive zero of a mixed Bessel function

The present article is concerned with lower and upper bounds of the first positive zero of the function H_ν(z, α) = αJ_ν(z) + zJ′_ν(z), Where J_gn(z) is the ordinary Bessel function of order ν > −1 and J′_ν(z) is the derivative of J_ν(z). A lower bound found here improves and extends the range of validity of the order ν, of a lower bound found in a previous work [8]. Also, two upper bounds given here improve a previously known upper bound [8]. In the particular case α = 0, these bounds lead to lower and upper bounds for the first positive zero j′_ν,1 of J′_ν(z) which improve well-known bounds in the literature.     

Variables sampling plans with PPM fraction of defectives and process loss consideration

Acceptance sampling plans provide the vendor and the buyer decision rules for lot sentencing to meet their product quality needs. A problem the quality practitioners have to deal with is the determination of the critical acceptance values and inspection sample sizes that provide the desired levels of protection to both vendors and buyers. As today's modern quality improvement philosophy, reduction of variation from the target value is the guiding principle as well as reducing the fraction of defectives. The C_pm index adopts the concept of product loss, which distinguishes the product quality by setting increased penalty to products deviating from the target. In this paper, a variables sampling plan based on C_pm index is proposed to handle processes requiring very low parts per million (PPM) fraction of defectives with process loss consideration. We develop an effective method for obtaining the required sample sizes n and the critical acceptance value C₀ by solving simultaneously two nonlinear equations. Based on the designed sampling plan, the practitioners can determine the number of production items to be sampled for inspection and the corresponding critical acceptance value for lot sentencing.     

Production quality and yield assurance for processes with multiple independent characteristics

Process capability indices have been widely used in the manufacturing industry providing numerical measures on process potential and process performance. Capability measure for processes with single characteristic has been investigated extensively, but is comparatively neglected for processes with multiple characteristics. In real applications, a process often has multiple characteristics with each having different specifications. Singhal [Singhal, S.C., 1990. A new chart for analyzing multiprocess performance. Quality Engineering 2 (4), 397–390] proposed a multi-process performance analysis chart (MPPAC) for analyzing the performance of multi-process product. Using the same technique, several MPPACs have been developed for monitoring processes with multiple independent characteristics. Unfortunately, those MPPACs ignore sampling errors, and consequently the resulting capability measures and groupings are unreliable. In this paper, we propose a reliable approach to convert the estimated index values to the lower confidence bounds, then plot the corresponding lower confidence bounds on the MPPAC. The lower confidence bound not only gives us a clue minimum actual performance which is tightly related to the fraction of non-conforming units, but is also useful in making decisions for capability testing. A case study of a dual-fiber tip process is presented to demonstrate how the proposed approach can be applied to in-plant applications.     

Capability-based quick switching sampling system for lot disposition

Various acceptance sampling schemes have been developed for quality control and assurance. In this research, two types of variables quick switching sampling (VQSS) system based on the process capability index C_pk are proposed. The VQSS is composed of two single sampling plans, one is under a normal inspection and the other is under a tightened inspection. Requirements for accepting a lot under the tightened inspection are more stringent than under the normal inspection. The concept of the VQSS system is that the sampling mechanism can adjust flexibly based on the quality history of the preceding submitted lots. A minimization model is constructed to solve the plan parameters for each type of the VQSS system under different mixes of quality levels and risk endurance levels, and several tables are prepared for references. In addition, the performance of the two types of VQSS system are compared with the single sampling plan through the operating characteristic (OC) curve and the average sample number (ASN) required for inspection. Finally, a real example from a dielectric layer coating machine is presented to show the practicality of the proposed system.     

Tool replacement policy for one-sided processes with low fraction defective

In most manufacturing industries, tool replacement policy is essential for minimizing the fraction defective and the manufacturing cost. Tool wear is caused by the action of sliding chips in the shear zone, and the friction generated between the tool flank and workpiece. This wear, apparently, is a dominant and irremovable component of variability in many machining processes, which is a systematic assignable cause. As the tool wear occurs in the machining processes, the fraction of defectives would gradually become significant. When the fraction defective reaches a certain level, the tool must be replaced. Therefore, detecting suitable time for tool replacement operation becomes essential. In this paper, we present an analytical approach for unilateral processes based on the one-sided process capability index C _PU (or C _PL ) to find the appropriate time for tool replacement. Accurate process capability must be calculated, particularly, when the data contains assignable cause variation. By calculating the index C _PU (or C _PL ) in a dynamical environment, we propose estimators of C _PU (or C _PL ) and obtain exact form of the sampling distribution in the presence of systematic assignable cause. The proposed procedure is then applied to a real manufacturing process involving tool wear problem, to demonstrate the applicability of the proposed approach.     

Capability adjustment for gamma processes with mean shift consideration in implementing Six Sigma program

In the 1980s, Motorola, Inc. introduced its Six Sigma quality program to the world. Some quality practitioners questioned why the Six Sigma advocates claim it is necessary to add a 1.5σ shift to the process mean when estimating process capability. Bothe [Bothe, D.R., 2002. Statistical reason for the 1.5σ shift. Quality Engineering 14 (3), 479–487] provided a statistical reason for considering such a shift in the process mean for normal processes. In this paper, we consider gamma processes which cover a wide class of applications. For fixed sample size n, the detection power of the control chart can be computed. For small process mean shifts, it is beyond the control chart detection power, which results in overestimating process capability. To resolve the problem, we first examine Bothe’s approach and find the detection power is less than 0.5 when data comes from gamma distribution, showing that Bothe’s adjustments are inadequate when we have gamma processes. We then calculate adjustments under various sample sizes n and gamma parameter N, with power fixed to 0.5. At the end, we adjust the formula of process capability to accommodate those shifts which can not be detected. Consequently, our adjustments provide much more accurate capability calculation for gamma processes. For illustration purpose, an application example is presented.