The space allocation problem (SAP) is the process in architecture, or in any kind of space planning (SP) technique, of determining the position and size of several elements according to the specified requirements. These are usually topological and geometric constraints, as well as matters related to the positioning of openings according to their geometric dimensions, in a two- or three-dimensional space. This was a traditionally challenging computable problem due to the combinatorial complexity of many different possible arrangements of a variety of architectural elements. Machine learning techniques represent an important set of tools for advancing solutions.
Variations Different compositions of the space being optimized for rooms and objects lead to various variations which have their own unique challenges. A one dimensional problem (ODSAP) might mean arranging rooms along a corridor, which becomes an ordering problem, attempting to determine the optimal ordering of rooms based on trips between those rooms minimizing the distance traveled by denizens of the building. The 'Multi-story Space Allocation Problem' (MSAP) deals with buildings with multiple floors which might involve vertical interconnections like HVAC or elevators.
References
