We rely on our homes to be bastions of comfort (or at least security and safety) in adverse conditions as well as fair, but their design frequently fails to keep this promise on many levels. For example, in climates where extended sub-freezing temperatures are the norm, we build plumbing susceptible to extensive damage if indoor heat is not maintained, yet have heating rely on grid services prone to interruption. Or we (re-)build homes in flood zones, or at forest-edges where wildfires rampage, etc.
For any given individual site, it may be difficult to construct an accurate cost model of non-resiliency, but collectively we know the environmental burden is significant. The reactive house attempts to determine and mitigate known risks, to insure that houses are protected against damage and ill-function. These risks include short term, but also (to the best of our ability to forecast) longer term issues such as more violent storms, heat extremes, rainfall, technology change, etc.
Resilience measures are not confined to climate/weather events: we should also consider changes in occupancy, or inhabitant behaviors/challenges. For example, how will a house need to be modified to become wheelchair accessible? So-called universal design is a good example of a resilience-promoting practice.
However, getting resilience right can be very difficult - since our ability to model the types of disruption in the future is quite limited. Consider attempts to 'future-proof' domestic spaces: not long ago, it was widely recommended we incorporate a vast network of cables running to every room in the house (Coax, Cat-5,6, phone, etc) to anticipate the delivery of digital services. What a waste of resources, labor, etc when wireless technology swept most competitors aside. Utilizing such frameworks as 'Open Building' can minimize resilence mis-predictions.