Food4Rhino

iGene is developed by the SD-II Lab (Sustainable Design and Intelligent Inference) at Tongji University to enhance the efficiency of intelligent optimization in performance-based design.

The plugin enhances data transfer efficiency, significantly improving the optimization performance of genetic algorithms when using surrogate models. It addresses a common bottleneck in the Grasshopper environment—long model loading times for every prediction.

Users can run any machine learning-based surrogate model for optimization tasks and fully benefit from the acceleration such models provide. Beyond surrogate-assisted optimization, the plugin also supports standard optimization loops, eliminating the need for manual intervention during multi-iteration tasks.

We provide several examples: one for general optimization and another showcasing surrogate model-based optimization. In addition, the plugin includes a pre-trained surrogate model and multiple use-case demonstrations to help users quickly implement and experiment with various optimization workflows. Notably, iGene performs exceptionally well across both general and surrogate-driven optimization scenarios.

If you encounter any issues, you can provide feedback to the SD-II lab at tongjisdiilab@gmail.com or to the developer within the team at zhong.plusplus@gmail.com.

Papers published with iGene for reference:

[1] Wang, Jinyu, Zhong, Zhengjia, Huang, Chenyu, Yao, Jiawei*, Zhang, Xiaoyu. iGene Plugin: A Novel Optimization Framework for Machine Learning Surrogate Model Integrated Performance-Driven Design[C]. Architectural Informatics, 2025, 30(1): 131–140. (CAADRIA 2025 Best Paper Award Runner-up)

[2] Li R, Huang C, Xin W, Ye J, Zhang X, Qu R, Wang J, Yuan L, Yao J*. Data-driven optimization reveals the impact of Urban Heat Island effect on the retrofit potential of building envelopes[J]. Building and Environment, 2025, 269: 112367. (SCI Q1, IF 7.1)