Complexity is an important element in axiomatic design theory. The current method for calculating complexity for a system following normal distribution is unbounded and approximate. The purpose of this paper is to present a detailed bounded solution for complexity using design and system ranges on a single function requirement.
This paper discusses the complexity measure for a system following a uniform distribution. The complexities of two types of systems, a system performing with a uniform distribution and a system performing on target according to a normal distribution are then considered and compared. The research proposes a complexity measure for a system performing within specification limits with a uniform distribution. In addition, a new concept of relative complexity is proposed.
A bounded solution for complexity for a normal distribution based on the existing assumptions was given which includes bias in addition to variance. The bounded solution was then compared to the existing approximate solution from the variance as well as bias standpoint. It was found that bias has an inappropriately reverse relationship with the bounded solution of complexity. Therefore, complexity cannot be used to approximate the system improvement when the improvement is based on a reduction in bias.
The current method for calculating complexity for a system following normal distribution is unbounded and approximate. This paper proposed a complexity measure for a system performing within specification limits with a uniform distribution.
Sharma, N.K. and Cudney, E.A. (2015), "Application and limitation of axiomatic design complexity in quality engineering", International Journal of Quality & Reliability Management, Vol. 32 No. 1, pp. 3-17. https://doi.org/10.1108/IJQRM-04-2013-0071
Emerald Group Publishing Limited
Copyright © 2015, Emerald Group Publishing Limited