Understanding consciousness is one of the greatest scientific challenges of this century. The Neurosciences Institute undertook a project to use the neurorobotic methodology to model the neural basis of consciousness. Given the strong link between consciousness and mental imagery, McKinstry focused on a classic mental imagery task -- mental rotation. Building on previous work with large-scale, spiking-neuron models of cortical networks, he constructed a model that controlled a humanoid robot that could learn to solve this task. The model showed behavioral and neural signatures that, in humans, are taken as evidence for mental imagery.  He will describe the model architecture, results, and challenges of the project as well as future directions for building a conscious artifact.