# U-Net-Accelerated Quality-Diversity Optimization for Climate-Adaptive Urban Layouts

arXiv: https://arxiv.org/abs/2606.04658 · HTML: https://mlsanity.com/paper/2606.04658/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 20 of its 37 references that are in the map.
- First posted on arXiv 3 Jun 2026.

## Builds on (most cited first, 20 of 20)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Image-to-Image Translation with Conditional Adversarial Networks](https://mlsanity.com/paper/1611.07004/) (2016)
- [GPyTorch: Blackbox Matrix-Matrix Gaussian Process Inference with GPU Acceleration](https://mlsanity.com/paper/1809.11165/) (2018)
- [Machine learning accelerated computational fluid dynamics](https://mlsanity.com/paper/2102.01010/) (2021)
- [Illuminating search spaces by mapping elites](https://mlsanity.com/paper/1504.04909/) (2015)
- [Scalable Variational Gaussian Process Classification](https://mlsanity.com/paper/1411.2005/) (2014)
- [Deep Learning Methods for Reynolds-Averaged Navier-Stokes Simulations of Airfoil Flows](https://mlsanity.com/paper/1810.08217/) (2018)
- [Covariance Matrix Adaptation for the Rapid Illumination of Behavior Space](https://mlsanity.com/paper/1912.02400/) (2019)
- [Differentiable Quality Diversity](https://mlsanity.com/paper/2106.03894/) (2021)
- [Using Centroidal Voronoi Tessellations to Scale Up the Multi-dimensional Archive of Phenotypic Elites Algorithm](https://mlsanity.com/paper/1610.05729/) (2016)
- [Quality-Diversity Optimization: a novel branch of stochastic optimization](https://mlsanity.com/paper/2012.04322/) (2020)
- [Deep Surrogate Assisted Generation of Environments](https://mlsanity.com/paper/2206.04199/) (2022)
- [Data-Efficient Design Exploration through Surrogate-Assisted Illumination](https://mlsanity.com/paper/1806.05865/) (2018)
- [pyribs: A Bare-Bones Python Library for Quality Diversity Optimization](https://mlsanity.com/paper/2303.00191/) (2023)
- [Deep Surrogate Assisted MAP-Elites for Automated Hearthstone Deckbuilding](https://mlsanity.com/paper/2112.03534/) (2021)
- [Model-Based Quality-Diversity Search for Efficient Robot Learning](https://mlsanity.com/paper/2008.04589/) (2020)
- [Uncertain Quality-Diversity: Evaluation methodology and new methods for Quality-Diversity in Uncertain Domains](https://mlsanity.com/paper/2302.00463/) (2023)
- [Designing Air Flow with Surrogate-assisted Phenotypic Niching](https://mlsanity.com/paper/2105.04256/) (2021)
- [From large-eddy simulations to deep learning: A U-net model for fast urban canopy flow predictions](https://mlsanity.com/paper/2507.06533/) (2025)
- [Full Domain Analysis in Fluid Dynamics](https://mlsanity.com/paper/2505.22275/) (2025)

## Abstract (arXiv)

Optimizing urban layouts for climate adaptation requires balancing building density with cold-air ventilation. Because physics-based climate simulations are computationally expensive, planners typically evaluate fewer than ten manual designs. glsqd algorithms offer a way to systematically illuminate the design space, but they require surrogate models to be practical. In this paper, we replace a slow, regulatory physics simulator with a spatial deep-learning surrogate (U-Net) inside an offline MAP-Elites loop. We systematically compare this spatial approach with a traditional glsgp surrogate across different training-data strategies (quasi-random Sobol sampling vs. active glsqd bootstrapping). Our results reveal that scalar glsgp surrogates fail catastrophically when trained on random samples, requiring expensive, actively generated glsqd archives to generalize. In contrast, the spatial inductive bias of the U-Net allows it to learn the underlying physics mapping robustly (R2 = 0.996), completely independent of the training data source. This allows offline glsqd optimization to achieve highly accurate fitness rankings (ρ = 0.994) using only a one-time batch of random training samples. The resulting pipeline, deployed in the open-source OpenSKIZZE tool, generates thousands of diverse, climate-evaluated building layouts in under ten minutes.