Engineers like the phrase "design spiral", to refer to a development process characterised by a series of iterative improvements and additions, particularly where they may be needed to meet evolving requirements. The term comes from software design, and is seen as a dramatic improvement on the linear design process which usually characterises more traditional manufactured products.
What makes the spiral attractive is the potential to continuously renew a system so that it always incorporates the latest technologies and meets the latest needs. In a linear design process, specified requirements would be followed by design, prototyping, testing and manufacturing. Each step is completed before work flows over to the next stage, and hence the process is sometimes referred as the "waterfall" design method.
Big aircraft programmes which begin with a request for proposals that lead to carefully detailed performance specifications are good examples of the waterfall approach. Given the enormous investment in tooling and materials required to make such a complex physical (as opposed to software) system, this linear approach seems unavoidable. However, because it's easy to fall off a waterfall but very difficult to go back up, adding new technologies or changing components is very expensive once an aircraft is in production.
To cite a specific and very current example, one of the factors contributing to Boeing's travails with its much-delayed 787 airliner programme is arguably what amounted to an attempt at highly compressed spiral design, by carrying out the prototyping, testing and manufacturing phases simultaneously. Looked at from that perspective, the 787 would illustrate why in popular language the word "spiral" is so often associated with the phrase "out of control".
But the language made a small spiral forward this week at the DSEi defence equipment exhibition in London. Gianfranco Terrando, senior vice president for unmanned air systems at Finmeccanica's Selex Galileo unit, referred to a flight test planned for later this year on a General Atomics Predator B of a new open architecture concept in payload integration as "spiral zero".
The concept is to create a family of options for packaging sensors and data management software so that customers can, ultimately, combine any sensors and data handling techniques with whatever airframe they like. A spiral development process makes such a concept possible - if each new requirement demanded that engineers go back to stage one and devise a system from scratch, the cost and time involved would, clearly be prohibitive.

Recent Comments