Day three brings in Galapagos and genetic algorithms. We start with what optimization actually is: conventional versus parametric design, fitness landscapes, when to use it and when not to, and the Galapagos hyperparameters that matter (population size and initial boost). Then we put it to work on three projects, from finding the smallest bounding box for freeform facade panels to tuning an attractor-driven opacity facade to a target percentage. The day ends with a cylindrical tower with a wavy NURBS facade and angle-grouped zigzag panels, where four parameters drive a multi-objective optimization with Ladybug sun-hour analysis and EPW data to balance over-exposed and under-exposed cells.
source



