Fixation on premature concept choices - a pitfall of early prototyping?

Abstract Building prototypes is an essential element in conceptual design. We argue that using resources to build prototypes may induce adherence to the chosen concept and prevents further exploration into other concepts. This phenomenon has previously been attributed to sunk cost caused by building prototypes. In a controlled human subject experimental study in a robot development context we investigate the influence of building and testing prototypes by allowing one group of participants to test their prototypes frequently while the other group is not allowed to test and can only rely on the provided information about the hardware. We report about participants prematurely committing to concept choices and adhering to those after building and testing prototypes while non-testing participants make superior concept choices based on the provided information. While planning may be feasible in some projects with low uncertainty, problems that are more complex require prototyping for knowledge acquisition. We give suggestions on how to reduce the costs of prototyping and the associated effect it has on design fixation. These suggestions are very similar to the test driven development approach known from software development. They include the definition of critical functions and the respective tests before building the prototypes. When designing the prototypes the focus lies on making a conscious choice of how to prototype with the lowest fidelity necessary to comply with the previously defined test and attempting risky development stages early with the intention of maximizing the work not done.