User talk:Systonomy

Design For Six Sigma in IT & Software Engineering
Six Sigma has been extremely successful when applied to manufacturing and transactional processes and large efficiency gains and savings have been made in those industries. However, Six Sigma remains almost unknown in the fields of Information Technologies (IT) and software engineering despite the obvious need for radical improvements in these industries.

Why is this? Do we assume IT and software engineering to be so different from other industrial processes that Six Sigma methods don’t work?

Misconceptions & Considerations
The main issue would seem to be the notion that software engineering is “intangible” and “innovative” of nature and is therefore not suitable for the application of rigid disciplines such as Six Sigma. Unfortunately this notion has been reinforced by several companies who have attempted to apply manufacturing or transactional versions of Six Sigma without adaptation.

While Six Sigma concepts and methods are generally useable across a broad range of industries it has always been known that adaptation for specific industries would be required. Therefore we must accept that software development is not primarily a manufacturing activity. It is an engineering project management activity and therefore there are technical aspects that need to be understood in Six Sigma terms:

• Software Engineering is a social discipline; it is a mixture of technical, organisational and social knowledge. Unlike other engineering disciplines, there is no separation between the knowledge of how to develop products and the knowledge of how to organise development processes. Often software processes are macro (or even meta) processes that rely on individual skills and competences; process variation can never be eliminated.

• Software engineering is a particular engineering discipline where the work is mostly on models and rarely on real world objects. All deliverables from requirements, to architecture, to design are just models, including the final product itself (i.e. Software or an Information System); it is a representation of a real world situation. The quality of the final product lies in the modelling power and techniques used to express the problem.

• Information and Software Systems are layered systems. They are built upon systems, which are themselves built upon other systems. They are designated as complex systems having emergent properties (e.g. safety, security and some aspects of reliability).

• Last but not least, Information System and Software Engineering is a first and foremost an Engineering of Evolution and not only and Engineering of Construction. In that sense, it would be closer to life science and bioengineering than civil engineering and manufacturing. Though most of the analogies and terminologies come from civil engineering.

In addition to these inherent properties, there are many process improvement methods that are often perceived as competitors with Six Sigma such as CMM®, CMMI®, ISO Tickit, ITIL and more recently Agile methods. This often leads to misconceptions whereby organisations believe that Six Sigma is only applicable at high levels of maturity (e.g. CMM level 3 or 4) and not realising that Six Sigma is actually a continuous improvement method applicable at any maturity level.