The research investigates the development of complex architectural solutions that creatively engage with material organization, production and design performance as intrinsic properties to a creative design methodology, and opportunistically; capable of producing a range of aesthetic effects. Agent based programming offers an event-based approach to computer programming that is context sensitive and systemically open. As an object-orientated programming method, agents are essentially entities that are endowed with autonomy through preprogrammed decision making capabilities. Since Henri Poincarré’s 3-Body problem, it has been understood that where three or more discrete entities interact, a non-linear chaotic process ensues.

Craig Reynolds has demonstrated that emergent adaptive formations such as flocks of birds and schools of fish can be simulated through very simple agent-based vector functions. Scientists Eric Bonabeau and Guy Theraulaz have illustrated that termites and wasps utilize similar multi-agent protocols to construct their mounds and nests. Similar methods can and have been employed in individual and swarm robot systems for architectural design and construction purposes, with much potential for future development. 
At this stage, we developed an algorithmically determined tool paths for robot motion for 3d printing through bottom-up rule sets.  Here, we evaluated production constraints and attempt to incorporate these as heuristics within our tool path .