This just in: The latest paper by WPI professors Jamie Monat, PhD, and Thomas Gannon, PhD, is now available for download!
Systems, an international peer-reviewed journal on systems engineering, has published Monat and Gannon’s “Applying Systems Thinking to Engineering and Design.” Here’s the abstract from the journal:
The application of Systems Thinking principles to Systems Engineering is synergistic, resulting in superior systems, products, and designs. However, there is little practical information available in the literature that describes how this can be done. In this paper, we analyze 12 major Systems Engineering failures involving bridges, aircraft, submarines, water supplies, automobiles, skyscrapers, and corporations and recommend Systems Thinking principles, tools, and procedures that should be applied during the first few steps of the System Engineering design process to avoid such catastrophic Systems Engineering failures in the future.
Click here to download your copy
Congratulations to Profs. Gannon and Monat, and happy reading! Let us know what you think of the paper in the comments.
♥ Love this post? Subscribe to Catalyst
I am very happy to see a paper addressing Systems Thinking and Systems Engineering, and agree with most of your points.
However I think you do not go far enough and are missing a key thing (definition of properties of a system).
1) not far enough
I agree ST and SE are not the same – but I see ST as a core element of SE. A view of this is to see SE as the application of ST in an engineering domain using a systems approach / purpose. Some seem to get confused that because ST can be used outside engineering then SE must only be using a part of it. That is not true – the difference is the structure in which it is used and the domain. ST is a perspective – but more it is a “framework for curiosity” that enables the right questions to be asked – and the tools (of which there are a vast number) are simple ways of asking those questions, based around the properties of a system (see next point)
2) the key missing point is you don’t describe the properties of a system. to me the definition of Systems Thinking is applying properties seen in systems. These include – systems have a context, systems seek to achieve a purpose (natural systems are not designed, so don’t try to achieve, just do)m system purpose can be broken down in functions, systems are made of parts and exhibit properties of the whole not deducible from the parts in isolation (emergence), systems effect the environment, systems are effected by the environment, systems have a life cycle, systems exhibit dynamic behaviour (feedback, lags etc.), the understanding of a system depends on the perception of the observer (so different observers see different things), systems are made up of systems and are sub-systems of bigger systems, systems have structure, systems behaviour exhibit patterns (in time)
The tools are ways addressing those properties. I recommend the tools (all paper based) contained in http://www.burgehugheswalsh.co.uk (they provide SE training to RR and many others!)
It is important to kill the idea that it has to be a system to apply systems thinking. Anything can be considered as a system. the plural systems in systems thinking is the key – its about thinking in terms of the properties seen in a system.
I think describing it this way, and looking at the system property that was missed in each failure would be more helpful. I (and Burge hughes walsh) have used both millennium bridge and 20 French church street. I think some of your examples are possibly down to bad engineering not bad SE/ST. an important element I don’t think you discuss enough is the role of understanding the functions a system expected to perform, and the need to do Function failure mode and effect analysis – to see what happens if funcitons fail (which identifies functions to mitigate, contain, prevent and monitor functions)
The other aspect I think you don’t go into enough is the architecture. an important element (after N2) is to perform functional allocation from the system level functions to the system elements. This helps identify levels of interface, which is where systems break.
Related to I don’t agree with your overall process – it misses (steps 5 and 6) that systems are made of parts, and each part needs to follow a system design process, and then be integrated into the higher level system. Therefore after step 5 you need to system architecture, decomposing into parts, and then repeat of SE process for all parts (some concurrency) and a step to integrate after this – replacing “detail design”. as this stands it is in danger of becoming engineering a system and missing the critical point that systems made of sub-systems – which can be considered systems in their own right.
Happy to convers off-line. Look at any of my INCOSE papers and you will see this view repeated frequently!
Hi Richard—we welcome your feedback on our paper, and appreciate your perspective. Of course, our definition of Systems Thinking is not quite the same as yours, and this lack of consensus on the definition of Systems Thinking is well-known and a little troubling, especially for practitioners. In fact, we are planning on holding a colloquium to address this very issue in the autumn of 2019—hope you will be able to attend and even present a paper.