BIM
A good façade design takes various criteria into account. These are often in competition with each other. An optimisation of one criterion of a façade design (e.g. daylight admission) can hypothecate the performance of the façade on another criterion (thermal performance).
By designing the façade parametrically, the design space can be explored within the set constraints (the design requirements) by automatically generating and analysing a large number of alternatives. Machine Learning techniques can be used to ensure that the examined alternatives converge to a global optimum for the entire criteria set.