EXPLORING RESUMPTION AIDS FOR TASKS INVOLVING 3D MODELS ON TABLETS

Abstract

Aircraft mechanics use mobile devices and workstations to view 3D models of parts and assemblies. This offers easy viewing of static perspectives of a 3D model; however, during more complicated tasks when model perspectives need to be rotated, panned, and zoomed, mechanics can experience difficulty resuming the task after interruption, either due to not recalling details of the work, or not recognizing the current perspective in relation to the part or assembly as a whole. In an attempt to reduce the loss of task context after interruption, we developed three resumption aid interfaces. Each provides a different method of capturing work state at key moments for later retrieval. They are: screenshots with audio, video-only, and combined audio/video. We conducted a comparative evaluation (N=32) with four resumption conditions (one for each interface and a control condition with no resumption aid). Participants used the respective aid to record their work while performing tasks. Tasks were interrupted, and resumed in a second session. Our participants used the data recorded in the resumption aids for two purposes: at the beginning of the resumed task, to assist in locating the resumption point, and at the end of the resumed task to assist recalling a task’s steps for the purpose of writing a task summary. Findings show that these tools boost interruption recovery for tasks based on 3D models. Using a resumption aid to record work state in the first session led to faster resumption in the second session if participants did not reference the assigned resumption aid at the point of resumption. If they did use the tool as a reference, resumption time was comparable to the control condition (no resumption aid). When recalling task steps at the end of a task, accuracy was highest when the resumption aid was referenced. Participants preferred combined audio/video and screenshots with audio to the video-only interface. Our results support the memory for goal theory (Altman and Trafton, 2002), which asserts that supporting activation of the interrupted goal and encoding cues before the onset of the interruption will aid later resumption. Our results suggest that audio annotations anchored with screenshots or video help users recall subtasks when working with complex 3D models, and that this is helpful for both task resumption and summarization.