Scaling Up Wearable Cognitive Assistance for Assembly Tasks

<p> Wearable Cognitive Assistance (WCA) applications run on wearable mobile de-<br> vices, to provide guidance for real world tasks. Physical assembly tasks have been a<br> significant focus of research on WCA. We introduce new techniques to support the<br> development of WCA applications for more complex assembly tasks than previous<br> techniques supported. In addition, our work reduces the load on developers creating<br> WCA applications by eliminating the need to collect and label real training images.<br> We accomplish this by training computer vision models on synthetically generated<br> images. This dissertation investigates escalation to human experts in cases when a<br> user is not satisfied with the automated guidance from an application. Lastly, we<br> develop a new version of a software framework for WCA applications, and evaluate<br> ways in which WCA applications can benefit from running computations directly on<br> mobile devices. </p>