DESIGN POSSIBILITIES OF EMERGENT ALGORITHMS FOR ADAPTIVE LIGHTING SYSTEM

Abstract

Algorithm aided design methods in architectural design provide mechanisms
for the use of complex adaptive systems that are self-regulating,
bottom-up processes where global behaviour emerges from the interaction
of autonomous agents. Including other systems that display
emergent properties, these methods can be categorised as emergent
algorithms and they are capable of generating organic patterns, representing
the complexity and intricacy of natural systems and displaying
high levels of distributed intelligence.

This paper explores the use of different emergent algorithms, such as
Cellular Automata, Lindenmayer system and Swarming Algorithms in
relations to an adaptive lighting system that reacts to users and changes
in environmental conditions. The design of this kind of computational
model requires new design methods and tools for control in enabling
design exploration. A graph-based approach is presented as a possible
solution for the difference between the continuous virtual coordinate
space and the actual scarce placing of urban light sources. The light
sources and sensors are mapped as a network of interconnected nodes
and it is used in defining the topology of the virtual coordinate system.
The challenges defining the functionality and enabling design using the
emergent methods are explored and possible solutions are discussed.