Proceedings of the Annual Meeting of the Cognitive Science Society

Subject of this article is the question whether the potential forautomatic defect analysis for symbolic timed ACT-R models asdemonstrated in earlier work can be developed into a scalable andcomprehensible technique. We present a formal, operational modelof an ACT-R architecture and a translation scheme of ACT-Rmodels into timed automata. We have applied this translationto ACT-R models and report on scalability experiments withautomatic defect analysis.