At this point, you might think about the difference between planners and production systems. Apart from GPS, few planners have been used as models of human planning, whereas there are many production system models. Planners and production systems are both programs, but what is the difference between them?
First, the input --- the specification of the problem to be solved --- is in a rather different form. A planner is specialised for reading this specification as a list of actions and their effects (e.g. GPS' operator tables) plus a statement of the current world state and the goal. A production system (at least if used as a general model) must be much more general, because it may have to deal with many other kinds of input. In fact, to call the action and goal specifications input is a bit misleading, since they were probably created by the system in the first place. Indeed, it's possible that their creation and the act of planning might not be separable --- are they in people?
Second, a planner contains a lot of specialised knowledge about how to plan --- how to abstract, how to re-order goals, and so on. Where is this knowledge in a production system? Certainly not in the inference method.