MANIFESTO
Urban climate risk cannot be solved with qualitative slogans. We translate the resilience of green infrastructure into millimeters of runoff reduction and degrees of thermal-comfort gain, and replace days-long CFD runs with second-scale GNN surrogates. Our work is to make green infrastructure measurable.
- Founded
- 2021
- Core domains
- 06
- Offices
- Suwon · Seoul
DOMAINS
Four urban disaster domains we cover
We combine simulation, data and spatial optimization for each disaster type.
Cloudburst · Urban Flooding
SWMM/LID + Neural ODE validated across 21 storm events (2018–2023) over a 47 km² urban catchment, with dynamic AMC coupling and 33.8% peak-flow attenuation scenarios.
Urban Heat · Heat Island
OpenFOAM CFD + GATv2 surrogates (zero-shot R²=0.845), Bayesian-optimized street-tree spacing (5–8 m) delivering up to 1.98 °C block cooling.
Cold Snap · Wind Environment
Cold-exposure analysis in urban canyons, simulation of windbreak and buffer planting, wind-field distribution analysis.
Fine Dust · Air Quality
Quantifying canopy filtration and airflow dispersion, optimizing park and green-belt placement to reduce pedestrian exposure.
EXTENDED
Extended Capabilities — Beyond the City
Beyond the four urban disaster domains, we extend analysis to infrastructure safety and rural climate-crisis response.
Street-Tree Safety · Fatigue Life
Wind-load and decay-coupled fatigue life assessed by ABAQUS dynamic FE + Monte Carlo, with species- and climate-regional inspection thresholds (Korean Ministry of Environment R&D 2022003570004).
Rural Climate Crisis · AI & IoT Response
LSTM-based 6-hour runoff prediction for farm catchments, LID effect verification, and IoT/Kakao-talk alerts — buying decision time for pumps, drainage gates, and resident evacuation.
WORK
Featured Work
Selected green-infrastructure disaster-resilience analyses for public, private and R&D clients.
Seocho-gu urban hydrology — GE-Water + Neural ODE soil-moisture dynamics (47 km², 21 storm events)
Seocho/Gangnam dynamic-AMC LID simulation (2,497 conduits, 15 scenarios)
Seocho-gu permeable-pavement AI spatial optimization (XGBoost + NSGA-II + SHAP)
APPROACH
What sets us apart
Not just consulting — quantitative decision tools rooted in physics-based models.
Quantified resilience
Qualitative greening effects translated into mm of runoff reduction, °C of thermal-comfort gain, and years of fatigue life.
AI surrogate models
GATv2 GNNs replace days-long CFD runs with second-scale inference; LSTMs forecast 6-hour runoff in 0.01 s.
Spatial optimization
NSGA-II and Optuna propose Pareto-optimal green-infrastructure / LID layouts under budget and area constraints.
RESEARCH
Research & Insights
Six SCI(E) papers currently under peer review — our latest evidence base.
- 01
Continuous Soil Moisture Dynamics for Urban Stormwater Modeling Using Neural ODE: A Physics-Informed Approach for Seoul
Advances in Water Resources · Under Review
- 02
Dynamic Antecedent Moisture Coupling in Continuous EPA-SWMM Simulation: LID Performance Under Seoul Monsoon Variability (2019–2020)
Journal of Hydrology: Regional Studies · Under Review
- 03
AI-Enhanced Spatial Optimization of Permeable Pavement for Urban Stormwater Management: Surrogate Modeling, NSGA-II, and SHAP Analysis in Seocho-gu, Seoul
Climate Risk Management · Under Review