Multi-criteria decision making for collaborative design optimization of buildings

Purpose: The purpose of this paper is to improve current design processes by proposing a new approach based on multi-criteria optimization of the designed asset. Management of design in construction projects is a complex task since it involves collaboration between professionals in multiple disciplines. Traditionally, designers work with a single solution at a time which is iteratively modified according to the view points of all the consultants. This results in sub-optimal solutions. A multi-criteria approach is able to accommodate diverse view points of specialist consultants in a construction project, aiming at a better optimized building system. Design/methodology/approach: The shortcomings of current design practices are analyzed based on a literature review. It is found that current approaches involving single objective optimization or Pareto optimization are not adequate for supporting collaborative design processes. A new approach to managing multiple objectives in design is proposed. This involves performing multi-objective optimization, presenting a population of good solutions to the design consultants and selecting the best solution through an algorithm called RR-PARETO2 (Relaxed-Restricted Pareto) filtering. A software tool with a graphical user interface was developed. An example of the design of a building façade is taken to evaluate the application of this approach. Findings: The paper provides empirical evidence that a multi-objective optimization approach is able to provide support for the task of accommodating multiple viewpoints in design. The proposed methodology allows navigation through the solution space and pruning it visually by applying constraints. It is shown that the RR-PARETO2 is able to select a good compromise solution with the best trade-offs among all the objectives. Originality/value: The idea of visualizing and filtering a population of design solutions has been proposed by design researchers for a long time, but is not currently adopted in practice in construction projects. The idea of collaborative filtering of the solution space according to the viewpoints of all the consultants by visually applying constraints on design variables and objectives is a new concept, the ultimate aim being a better balanced built asset. This is the first time the RR-PARETO2 algorithm has been applied to building design.