Does architectural design optimization require multiple objectives? A critical analysis

This paper analyzes eight assumptions that underly the general consensus in the computer-aided architectural design community that multi-objective optimization is more appropriate for and more analogous to architectural design processes than single-objective optimization. The paper discusses whether (a) architectural problems are best formulated as multi-objective optimization problems, (b) architectural design optimization is only about negotiating tradeoffs, (c) multiple objectives require multi-objective optimization, (d) Pareto fronts represent design spaces, (e) Pareto fronts require multi-objective optimization, (f) multi-objective algorithms are efficient and robust, (g) evolutionary operators make multi-objective algorithms efficient and robust and whether (h) computational cost is negligible. The paper presents practical examples of combining multiple objectives into one and concludes with recommendations for when to use single- and multi-objective optimization, respectively, and directions for future research.