This paper describes a design process for custom wearable systems produced in an academic setting. A set of 245 wearable design defects from two distinct periods separated by six years in time is presented. These data identify aspects of the process which require significant developer effort. We show this effort using several views of the data, including time spent, design region affected, and distribution under an Orthogonal Defect Classification scheme.

A comparison of defect attribute distributions across the two separate design periods is given. The results show that growing interoperability requirements are increasing design complexity, and inducing greater debugging effort. In addition, the combination of increasing pin counts and decreasing physical dimensions produce a higher rate of critical defects (i.e., those that require major physical modifications before debugging can continue). Finally, although the nature of the defects themselves has changed with time, the mechanisms used to discover defects have remained relatively constant.