Key Points


  • Taking a multisided view of quality is essential to understand and satisfy an assortment of stated and unstated customer requirements and those of other stakeholders.

  • Generally, quality principles, systems, and methodologies applicable to manufactured products are equally valid in software. However, software products have their own design and development environments. The complexities associated with many software products must be understood, given the novelties and difficulties of the tasks they are often designed to handle.

  • The task of producing trustworthy software is truly challenging and calls for genuine management involvement.

  • Traditional quality control systems are based on two fallacies: First, customer requirements are met as long as the product is within the specification limits, and second, business implications of the quality being "barely within specification limits" and those of "right on target" are the same.

  • Deming's 14 Points for Management for improving quality include listening to the voice of the customer, reduction of variation, use of statistical measures, winning the confidence and respect of coworkers, and continual improvements. He pointed out the inadequacy of a quality system depending on inspection.

  • The Japanese industrial engineer Genichi Taguchi developed an alternative quality management system, Taguchi Methods. It emphasizes the value of "right on target" and addresses quality effectively upstream rather than depending on inspection to detect and correct faults downstream.

  • The Taguchi quality philosophy can be summarized as follows:

    • Continual quality improvements and cost reductions are necessary for business survival.

    • An important measurement of quality is the total loss generated by that product to the societythe quality loss function.

    • Build quality into the product or process by design using Statistical Design of Experiments.

    • The customer's loss due to poor quality is nonlinear and can be approximately quantified as proportional to the square of the deviation of the performance characteristic from its target.

    • Product performance variation can be reduced by examining the nonlinear effects of "control factors" on performance characteristics.

  • Robust Design using Taguchi Methods is carried out in three stages: System Design, Parameter Design, and Tolerance Design.

  • Trustworthy software meets a variety of spoken and unspoken customer needs and must be customized. In an enterprise software context, a trustworthy software meets, at least, the following requirements: reliability, safety, security, maintainability, and customer responsiveness.

  • The DFTS process is characterized by the following:

    • Genuine leadership commitment and a supportive infrastructure

    • The ability to identify spoken and unspoken requirements using QFD

    • Optimization of customer requirements deploying Taguchi Methods

    • Establishing a concurrent coding and testing practice

    • Use of redundant software if required

    • Deploying appropriate quality and planning tools such as TRIZ, Pugh, and FMEA

    • Use of innovative software development tools such as OOD




Design for Trustworthy Software. Tools, Techniques, and Methodology of Developing Robust Software
Design for Trustworthy Software: Tools, Techniques, and Methodology of Developing Robust Software
ISBN: 0131872508
EAN: 2147483647
Year: 2006
Pages: 394

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net