A cross-platform optimization system for comparative design exploration of competing concepts and strategies

This study addresses a critical gap in architectural performance-based design optimization by introducing a cross-platform system that enables comparative evaluation of competing concepts and strategies for early-stage design exploration, where conventional optimization tools often prove inadequate due to the single-model optimization approach and overemphasis on numerical improvement. While optimization has demonstrated value in design space exploration, existing methods struggle to support parallel exploration and meaningful comparison of design concepts or strategies, limiting their utility in design ideation and critical decision-making. The presented cross-platform system bridges these gaps by integrating parametric and generative design with a dedicated evaluation server to create a coherent workflow for multi-model optimization, parallel performance simulation, and unified design and data visualization. The system enables designers to effectively manage complex tasks of optimization associated with multiple generative models, define meaningful performance evaluation functions, and facilitate them to conduct comparative evaluation of results from multiple optimizations. Two case studies demonstrate the system's capacity to reveal performance trade-offs between alternative design strategies and provide critical insight for decision-making. This study contributes infrastructure for comparative optimization-based exploration and evaluation of competing concepts and strategies for early-stage design. Moreover, the development of system emphasizes a user-oriented tool implementation through research, which is aimed to tackle practical challenges in design optimization, performance evaluation, data analysis, and information extraction. Compared with relevant works, the developed system synergizes the capabilities and flexibility of parametric and generative design with server-based scalability, while its practical value is evidenced by successful deployment in real-life design scenarios.