PhD Dissertation, Stanford University, 1997
Department of Mechanical Engineering. Advisor: Mark Cutkosky.
The thesis develops a framework in which multiple specialized agents work together through a shared design pool. Three types of agents handle different aspects of the problem: Creators generate initial candidate solutions; Mutators refine and improve existing solutions through local modifications; and Combinors merge the best elements from different solutions to produce stronger hybrids.
The human designer is not outside the system. The human is part of it. Agents handle the computationally intensive search and optimization, while the human contributes qualitative judgments that the agents cannot model: spatial intuition, aesthetic preferences, manufacturing knowledge, and the ability to recognize when a solution "looks right" even before formal metrics confirm it.
The design pool serves as shared memory. All agents, including the human, deposit solutions into the pool and draw from it. This allows each agent to build on the work of others without requiring direct coordination or communication protocols between agents.
Experimental results (Section 7.4.3) demonstrated that a human working collaboratively with specialized agents produced better solutions than either the human or the agents working alone. The collaborative system consistently found higher-quality cable routings in less time than purely manual design or purely automated optimization.
The question of how humans and autonomous agents collaborate effectively remains as relevant today as it was in 1997.