Control system testbeds and toys: serendipitous or suspect?

What role do experimental control system studies take with respect to theory, education, and design? Perceived advances in control systems research can easily be offset by suspect experimental evaluations resulting from narrow investigations into specific problems. Likewise, attention to implementation details is unquestionably a necessary component of any control design, often to the extent of actual experimentation. In this brief paper we provide philosophical views on issues related to the impact of experimental work on the field of control systems, and on the relative value of "toy" projects, "laboratory scale" testbeds, and full-scale experimental apparatus. In so doing we represent the university viewpoint, where experimentation serves a dual purpose: research and education.