Saskia Baltrusch

194 Chapter 7 Reducing muscular effort and unloading the back when wearing an exoskeleton has been assessed by previous studies, reporting reductions in L5S1 peak moments of 15-20 % [9, 12, 14]. This is comparable with our reductions in generated work around the supported joints (18-25 %). Besides the effect of taking over muscular effort, a concern when using passive exoskeleton is increased muscular effort in the legs, which has been reported when wearing the HappyBack and the Bendezy [31] and when using the BNDR [32]. However, in the present study work generated around the knee joint did not show differences between the conditions, indicating that the loading of the legs on average did not increase. The reduction in muscle work is also supported by the effects on muscle activity. Muscle activity showed clear effects of wearing the exoskeleton. The SPEXOR device reduced back muscle activity by 10 % to 16 %, indicating that less muscular effort was needed to perform the lifting task when wearing the exoskeleton. Even though there is not necessarily a linear relation between muscle activity and metabolic cost [33], the reduction in muscle activity is surprisingly similar to the observed reduction of net metabolic cost. Our EMG results are comparable to previous studies that assessed the effect of lifting devices on back muscle activity during repetitive lifting [8, 19]. Some studies reported higher reductions in back muscle activity [11,14,16]. However, they assessed the effect of wearing a passive exoskeleton during static bending, in which a continuous support of the exoskeleton is provided. In a dynamic task, such as repetitive lifting, this is not the case and this explains why we did not find as high reductions in back muscle activity. Alemi et al. (2019) [10] reported a reduction of back muscle activity by 29% when using the VT-Lowe’s exoskeleton during symmetric lifting. Potential causes for the greater reduction in comparison to the present study are the higher loads lifted by the participants in the study of Alemi et al. (2019) [10] and unknown difference in support provided by the respective exoskeleton. Abdominal muscle activity did not change, indicating that participants did not have to activate their abdominals to overcome resistance of the exoskeleton. Still, the reduced back muscle activity demonstrates the potential of the SPEXOR exoskeleton to unload the low back by reducing back muscle activity and hence reducing metabolic cost.