Metabolic Break-Even and Assistance Duration in Wearable Walking Exoskeletons Three Walking Conditions, Paired Percentages, and Route Energy

Abstract

Walking assistance lowers the user’s metabolic expenditure only when its benefit exceeds the energetic burden of the wearable apparatus. This article examines normal walking, inactive-device walking, and assisted walking using treadmill measurements from autonomous and tethered ankle exoskeleton experiments. Net metabolic power was 3.82, 4.01, and 3.43 W kg⁻¹, respectively, in seven adults using an autonomous device, and 3.60, 4.03, and 3.16 W kg⁻¹ in ten women using pneumatic ankle assistance. For an inactive-device penalty h and an assistance saving a relative to inactive walking, complete-system saving is s = (1 + h)a − h. Ratios of the condition means give complete-system savings of 10.21% and 12.22%. At fixed walking demand, alternating assisted and inactive intervals yields break-even assisted-time fractions of 32.76% and 49.43%, respectively. These fractions are calculated from group condition means; intermittent routes were not measured in those experiments. Participant-level averaging requires a covariance term, and unequal route durations require separate accounting of net locomotor and gross metabolic energy. Excess transition energy raises the assisted-time requirement. Energetic equivalence additionally requires an acceptable penalty margin and paired confidence intervals. The analysis identifies the measurements and statistical contrasts needed to determine whether a complete wearable device lowers walking expenditure under continuous and intermittent assistance.