Modeling aperture passage affordances in ACT-R 3D

In this paper I present a model of aperture passage judgment (judgment of whether an agent can walk through aperture, rotating shoulders as needed) and performance (initiation and termination of shoulder rotation while walking through an aperture) in ACT-R 3D. The model is adapted from Somers (2016) and represents a first attempt to unify findings across multiple experiments with a single model. The cognitive model is embodied in a robotics simulator, with motor movement controlled directly by the cognitive model. The model exhibits an improved fit as compared to Somers (2016), in the same experiment. The same cognitive model also exhibits a reasonable fit in additional, exaggerated conditions.