Task complexity in exoskeleton setup and takedown: Procedural steps and usability problems as predictors of deployment performance

Occupational exoskeletons show promise in reducing physical strain in industrial work, yet their industrial adoption remains limited. For exoskeletons to be used when performing tasks in industrial settings, they first need to be set up, then fitted and donned, and finally doffed, disassembled and stored for future use. Exoskeleton setup and takedown procedures can significantly impact industry deployment, adoption, and use, yet we have limited knowledge of the complexities in setup and takedown procedures and the resulting deployment barriers. The goals of this study were to understand how task complexity (number of task steps, usability problems and part count) in setup and takedown of exoskeletons impact task completion time and success, and present barriers in deployment. Twenty nine participants completed setup (assembly and donning) and takedown (doffing and disassembly) of four exoskeletons. We measured task times, success rates, task complexity (task step counts, part counts and usability problems). Hierarchical task analysis and heuristic assessments were performed to assess task complexity. Setup tasks, especially assembly, took the most time and exhibited the highest failure rates, whereas takedown tasks were faster and more successful. Participant-level regression analyses (N = 397 observations) showed that number of procedural steps was the strongest predictor, accounting for 66.9% of variance in completion times (β = 0.374, p < .001), with each additional step associated with approximately 22 seconds longer completion time. Usability problems also significantly predicted completion time (R² = 0.305, p < .001), while part count showed no significant association (p = .133). Our findings highlight that task complexity impact setup procedures significantly and can present a major deployment barrier. Moreover, our findings add new knowledge that number of steps, not part count, is the primary predictor of performance in setup procedures, and that usability problems differentially affect assembly and donning tasks.