How Changes in Muscle Activity and Range of Motion Represent User Perceptions of Back-Support Exoskeleton Performance

While different objective metrics have been used in the literature to evaluate exoskeletons' performance, less research has examined how these metrics represent users' perception of exoskeleton performance. This study aimed to investigate the relationships between muscle activity and joint kinematics with users' perceptions of a soft back-support exoskeleton performance across multiple subjective factors, including comfort, biomechanical support, pain, mobility restriction, and overall experience, while using a soft back-support exoskeleton. The study also examined how users' overall exoskeleton selections align with their pReferences across these subjective factors. Ten non-disabled individuals performed trunk bending and weight-lifting while wearing different exoskeleton configurations (eight configurations of four different sizes and two strength levels). Lower back muscle activity as well as trunk and knee joint kinematics were recorded using wearable sensors. A similarity index was used to quantify the similarity between users' subjective feedback across different factors, as well as the similarity between subjective feedback and objective metrics. The results showed that users' perceptions of comfort, absence of pain, and mobility restriction were highly consistent, with similarity indices exceeding 90% between these factors. Participants preferred low-strength elastic bands for comfort (75%), absence of pain (80%), mobility restriction (75%), and overall experience (68%), while higher-strength bands were preferred for biomechanical support (70%). Their overall exoskeleton selections reflected an intermediate compromise between comfort- and support-oriented configurations. Similarity analyses showed that comfort, absence of pain, and mobility restriction were more strongly associated with trunk range of motion (p < 0.01). Muscle activity represented human perception of support to some extent, with median similarity indices of 73.7% for trunk bending and 63.6% for weight-lifting. Overall, this research laid the foundations for understanding how different biomechanical metrics align with user perceptions of the soft exoskeleton performance and highlighted that users consider a broader range of factors when selecting their preferred exoskeleton setting.