AI-driven Adaptive Architecture for Climate Resilience

Course Description

AI-driven Adaptive Architecture for Climate Resilience is a three-day, hands-on course focused on building an AI-assisted workflow that connects microclimate data, building simulation, and machine learning to improve climate resilience at both building and urban-canyon scales. Participants learn how to predict heat stress, wind discomfort, and flash-flood exposure, then use these insights to optimize façades, shading, and ventilation strategies. The course also covers closed-loop control approaches—ranging from rule-based automation to MPC-style controls—supported by sensing concepts, fail-safes, and decision-grade KPIs. By the end, participants will produce a KPI dashboard and a reproducible workflow that can be adapted to real projects.

Aim

To equip participants to build an end-to-end AI workflow that predicts heat, wind, and flash-flood stress, optimizes façades/shading/ventilation, and supports closed-loop controls and urban-canyon adaptations using decision-grade KPIs.

Application Details

This course is delivered through an integrated learning and experimentation environment designed for climate-resilient architectural workflows. It includes structured learning modules, guided practical labs, reusable templates for data preparation and feature extraction, and assessment checkpoints. Participants work with simulation outputs and ML surrogate models to compare design alternatives, generate control schedules, and summarize results through KPI dashboards suitable for technical and stakeholder decision-making.

Course Objectives

Participants will be able to:

• Define resilience KPIs and identify key microclimate drivers affecting buildings and streetscapes.

• Assemble site and weather data, run baseline simulations (EnergyPlus/Honeybee/URBANopt and CFD where applicable), and extract usable features.

• Train and validate fast ML surrogate models for heat, wind, and flood indicators to enable rapid iteration.

• Optimize façade, shading, and ventilation strategies and auto-generate control schedules for performance targets.

• Design closed-loop strategies using rule-based logic and MPC-style approaches, including sensing plans and fail-safe mechanisms.

• Evaluate urban-canyon interventions, stress-test scenarios (heatwave + storm), and deliver a KPI dashboard with a reproducible workflow.

Course Structure

Module 1: Microclimate Foundations, Data, and Baseline Simulation

• Resilience scope: heat stress, wind comfort/safety, and flash-flood risk

• Microclimate drivers and practical KPIs for building- and street-scale decisions

• Site and climate data assembly: weather files, terrain/land cover, urban morphology inputs

• Baseline performance modeling: EnergyPlus/Honeybee/URBANopt workflow and simulation outputs

• CFD fundamentals for wind and canyon effects (when and why it matters)

• Practical lab: build the baseline model, extract features, and define KPI targets

Module 2: ML Surrogates and Design Optimization

• Feature engineering from simulation outputs and microclimate data

• Training fast surrogate models for heat, wind, and flood indicators

• Validation approach and uncertainty awareness for decision use

• Optimization workflows: façade parameters, shading control logic, ventilation strategies

• Auto-generation of control schedules and rapid comparison of alternatives

• Practical lab: train surrogates, run optimization, and shortlist design/control options

Module 3: Closed-Loop Controls, Urban-Canyon Adaptation, and Decision Outputs

• Closed-loop control concepts: sensors, actuation, control constraints, fail-safes

• Rule-based controls vs MPC-style strategies for façades, ventilation, and shading

• Urban-canyon resilience strategies: shading networks, ventilation corridors, surface/material choices, drainage considerations

• Stress testing: combined heatwave and storm scenarios, robustness checks

• Decision-grade reporting: KPI dashboard, trade-offs, and reproducible workflow packaging

• Practical lab: finalize dashboard and compile a project-ready workflow summary

Who Should Enrol

• Architects, façade engineers, building services engineers, and urban designers/planners

• Building performance, ESG/resilience leads, and sustainability consultants

• CFD/modeling practitioners and data/ML engineers working in AEC

• Graduate students and researchers in architecture, civil/environmental engineering, and urban climate

• Municipal resilience/Smart City teams and climate adaptation/risk professionals