Capability Analysis – Definition

Cite this article as:"Capability Analysis – Definition," in The Business Professor, updated September 20, 2019, last accessed June 6, 2020, https://thebusinessprofessor.com/lesson/capability-analysis-definition/.

Back to: OPERATIONS, LOGISTICS, & SUPPLY CHAIN MANAGEMENT

Capability Analysis Definition

Capability analysis refers to a set of tools and calculations used in determining whether a system meets certain specification requirements. It is the evaluation of a system or production process that entails statistical analysis of how well the process meets the given requirements. A set of data on the system or process is required to carry out the calculation.

A Little More on What is Capability Analysis

The nature of a system or production process as to whether it meets a set of specification is evaluated through capability analysis. To determine the capability of a process or system, it is vital to generate data relating to the system for the purpose of capability analysis. Process capaulify analysis helps to identify hoe adherent a process is to specific requirements and how suitable the process is for the purpose it is used. This analysis also helps to evaluate the specification tolerance of a system to USL  or LSLS. USL (upper specification limit) and LSL means lower specification limit.

There are three steps in capability analysis, they are;

  • Planning for data collection
  • Data collection
  • Plotting and analysing results.

Every process has natural limits and some specification or requirements that it should align to. How well a system merges its natural limits and specifications is discovered through capability analysis.  Capability study often reveal the state of a system (process), the changes or adjustments that occur to the process. During a process capability analysis, some of the following actions may be required;

  • No action: this is when nothing is meant to be done about the process, especially when the process meets the specification limits.
  • Adjust specifications: when the requirements or specifications are unattainable, the specifications need to be changed or adjusted. In a situation like this, the client must be informed before the specifications are modified, adjusted or changed.
  • Reducing variability: reducing the variability of a process is often a complicated thing to achieve but reducing the variation of a process is possible. It entails scrapping the initial specifications and reworking the whole process more efficiently.

Other capability applications:

In capability analysis, there are other vital procedures, these are otherwise called capability applications. They include the following;

  • Creating a basis for the establishment of a variable control chart.
  • Generating a set of data for capability analysis through a control chart.
  • Evaluating the new equipment.
  • Reviewing specification tolerances to the inherent variability of the process of system.
  • Assigning better equipment for the job.
  • A routine check and performance audits.
  • Adjusting and modifying the process.
  • Examining the effects of the adjustments.

Identifying Characteristics

Identifying the peculiar characteristics of a process is important in capability analysis. There are some requirements that the identification of characteristics to be examined in a capability study. They are;

  • The characteristics must be vital to the quality of the process
  • The values of the characteristics must be adjustable.
  • The conditions of the characteristics that are to be measured must be clearly defined.
  • The characteristics should be controllable.

Identifying Specifications/Tolerances

In capability analysis, identifying specification and tolerances is important. Typically, specifications and tolerances are influenced by the standards of the industry and the requirements of the customers. Other factors that determine the specifications and tolerances of a process the decision of the company’s engineering department and the company’s board.

Here are some key points you should note;

  • Capability study reflects how a system or process meets certain specifications while variation expresses how the process reflects certain tolerance requirements.
  • If a process is incapable as shown by the process capability study, information derived from the capability analysis helps to adjust the process to meet specifications and tolerance requirements. This means the specifications must be modified.

References for Capability Analysis

http://www.businessdictionary.com/definition/capability-analysis.html

http://www.statgraphics.com/process-capability-analysis

https://www.jmp.com/support/help/14/capability-analysis.shtml

https://www.greycampus.com/opencampus/lean-six-sigma-green-belt/capability-analysis

Academic Research for Capability Analysis

A methodology for probabilistic simultaneous transfer capability analysis, Xia, F., & Meliopoulos, A. S. (1996). IEEE Transactions on Power systems, 11(3), 1269-1278. This article presents the capability analysis of simultaneous transfer method on the basis of a probabilistic approach. In a power system of a large scale, all areas are categorized as (1) study area (2) transfer participating area (3) external areas that do not have any direct transactions. The authors apply a selection procedure of performance-index based contingency in the transfer and study areas for ranking those contingencies that have an effect on simultaneous transfer capability. A Wind Chime diagram variation utilizes this ranking order to specific contingencies. Then, an optimal power flow diagram evaluates them. Subsequently, the authors calculate the probability distribution of STC (Simultaneous Transfer Capability) on the basis of Markov models of electric load, circuit & unit outage.

A review of methods for measurement systems capability analysis, Burdick, R. K., Borror, C. M., & Montgomery, D. C. (2003). Journal of Quality Technology, 35(4), 342-354. This paper is a review of methods to analyze and conduct measurement systems capability researches emphasizing on the analysis of variance model. These researches are experiments performed using nested and crossed factors. The variance analysis is a captivating method to analyze the outcomes of the experiments since it allows efficient interval and point estimation of the variance elements linked to the variability source in the experiment. The authors show calculations for the standard 2-factor design, discuss designing experiment aspects and give references for the case in which this design is not applicable.

Recent developments in process capability analysis, Rodriguez, R. N. (1992). Journal of Quality Technology, 24(4), 176-187. This is an introductory paper published in the Journal of Quality Technology to address a special issue of the process capability analysis. The author outlines the statistical problems created using the current process capability indices. He highlights the contributions of the papers written on this problem and discusses what role the related graphical and statistical methods play in resolving the issues of the analysis.

Process capability analysis for an entire product, Chen, K. S., Huang, M. L., & Li, R. K. (2001). International Journal of Production Research, 39(17), 4077-4087. This article presents a flow path to examine an entire product’s process capability made up of several process characteristics. The flow path contains 6 steps. The flow path is applicable whether the process data follows a normal or non-normal distribution. The objective of this paper is to develop the MPCAC (Multi-Process Capability Analysis Chart) Model on the basis of Cpl, Cpn and Cpu. In a normal distribution, it examines the process capability. Similarly, this paper develops NMPCAC (Non-Normal Multi-Process Capability Analysis Chart) on the basis of Npu, Npn and Npl. In a non-normal distribution, it examines the process capability.

Process capability analysis—a robustness study, English, J. R., & Taylor, G. D. (1993). The international journal of production research, 31(7), 1621-1635. This paper analyzes the robustness of 2 common process capability ratios, Cpk and Cp when the random process under observation departs from normality. The authors derive the distributions of EPCR (Estimated Process Capability Ratios)and use them based on the validation of large-scale simulation researches in an evaluation of departure from normality. For recommended procedures, the analytical results and simulation researches prove to be a base. The authors suggest considering the effect of process distributions before using common process capability indices because of lack of robustness on departing from normality. The authors also use the function of the Taguchi loss to process capability analysis irrespective of the population distribution.

Working capability analysis of Stewart platforms, Luh, C. M., Adkins, F. A., Haug, E. J., & Qiu, C. C. (1996). Journal of Mechanical Design, 118(2), 220-227. This research analyzes the working capability of the platforms of Planar & Spatial Stewart on actuator length having unilateral constraints. The authors also examine the limits on the achievable motion at single configurations. They are linked to points interior to the output sets. As the working point moves on a spatial Stewart platform in 3-dimensional space, the output set’s boundary is a 2-dimensional surface. The used numerical approaches map 1-dimensional solution sets allowing the boundary to be featured by 1-dimensional generators. Similarly, the interior singular curves, inside the output set, characterize the motion control limit and analyze the barriers.

Process capability analysis of non-normal process data using the Burr XII distribution, Liu, P. H., & Chen, F. L. (2006). The International Journal of Advanced Manufacturing Technology, 27(9-10), 975-984. This research offers a novel alteration in the Clements’ method with the help of the Burr XII distribution to make the accuracy of indices estimates better linked to specification limits (only one-sided) for data of the non-normal process. The authors present a novel Burr-based technique and make its comparison with Clements’ method using simulation. Lastly, they propose an example application to semiconductor manufacturing and share the results of their process capability analysis.

Transmission interchange capabilityanalysis by computer, Landgren, G. L., Terhune, H. L., & Angel, R. K. (1972). IEEE Transactions on Power Apparatus and Systems, (6), 2405-2414. This paper develops a program to determine the interchange capability of the transmission system and those elements and outages which can limit the energy transfer in the interconnected power systems. The authors have tested the method extensively during the research on interchange capability of the MAIN (Mid-American Interpool Network). The program proves to be an operating tool and effective planning, minimizing the engineering and time required by the computer to evaluate interchange limitations.

Process capability analysis using personal computers, Pyzdek, T. (1992). Quality Engineering, 4(3), 419-440. This paper elaborates a few empirical and practical prospects of PCA (Process Capability Analysis). The purpose of this paper is to make a comparison between a controlled process and established requirements. As a result, the producer can estimate the long-term expected yields. One can use the PCA as a benchmark to estimate quality improvement. It also states the assumptions of this analysis and covers the PPIs and PCIs issues as well. The modern PCs have effectively solved the associated issues, such as the adjustment of PCIs for non-normality, missing data compensation, Curve fitting, off-centre target values, etc. The implementation of the PCA is performed using a commercial computer software package which the author developed.

A new approach to analysing non-normal quality data with application to process capability analysis, Shore, H. (1998). International Journal of Production Research, 36(7), 1917-1933. This study proposes a new method to determine non-normal quality data and demonstrates for PCA (Process Capability Analysis). This approach uses new distributions and its attached fitting procedures to estimate the distribution of an unknown source. As the new fitted routines need sample moments of low degree only, the fitted distribution is attached to appreciably smaller MSEs (Mean Squared Errors). The author develops 1st, 2nd and 3rd generation PCIs (Process Capability Indices) for non-normal populations, similar to present PCI designed for Pearsonian populations. For new PCIs, he compares sample estimators in terms of MSEs using Monte Carlo Simulation to existing 4-moment estimators.

Was this article helpful?