Tradeoffs between performance, cost and risk frequently arise during analysis and design of complex systems. Many such systems have both human and technological components and can be described by mathematical input-output models. Oftentimes such systems have known or desired outcomes or behaviors. This paper proposes “isoperformance” as an alternative approach for analyzing and designing systems by working backwards from a set of desired performance targets to a set of acceptable solutions. This is in contrast to the traditional “forward” process, which starts first in the design space and attempts to predict performance in objective space. Isoperformance can quantify and visualize the tradeoffs between determinants (independent design variables) of a known or desired outcome. For deterministic systems, performance invariant contours can be computed using sensitivity analysis and contour following. In the case of stochastic systems, the isoperformance curves can be obtained by regression analysis, given a statistically representative data set. Examples from opto-mechanical systems design and human factors are presented to illustrate specific applications of the method.
[1]
H. Cook,et al.
Product management : value, quality, cost, price, profits, and organization
,
1997
.
[2]
Achille Messac,et al.
Physical programming - Effective optimization for computational design
,
1996
.
[3]
Marshall B. Jones,et al.
Isoperformance Curves in Applied Psychology
,
1996,
Hum. Factors.
[4]
Robert S. Kennedy,et al.
A Meta-Model for Systems Development Through Life Cycle Phases: Coupling the Isoperformance Methodology with Utility Analysis
,
1990
.
[5]
Marshall B. Jones,et al.
Isoperformance and Personnel Decisions
,
2000,
Hum. Factors.
[6]
De Weck,et al.
Multivariable isoperformance methodology for precision opto-mechanical systems
,
2002
.