Here we see a fully autonomous robot by the name of "Cindy", navigating through a building, when a remote human operator gets in touch with her, informing her that Commander Z needs a medical kit and that one should be located in the room she is in. Cindy is able to infer the implicit instruction from the dialogue and offers her help to get the medical kit. When her offer is accepted, she forms a goal to look for a medical kit in the current room, but she also discovers that she does not know what the medkit looks like, so she asks and gets a description. From that description, she is able to build a visual model of the medkit, which allows her to look for it. Once she finds it on the table, she approaches the table and aligns herself with it, carefully looking for the handle which will allow her to pick it up. Although she has never picked up a medkit before, she is able to plan (through internal simulations) a trajectory for her right hand to grab the medkit, at which point she forms a goal to drive to the room with the green door, where she was told Commander Z is located. All reasoning, inference, natural language understanding, as well as perceptual and action learning algorithms are general and implemented in our DIARC control architecture.