“…EMG is a great tool for research but is difficult to use as the basis for accessible, scalable, practical ergonomic assessment, particularly for the back. To briefly summarize key challenges: (i) EMG requires considerable time and expertise to collect and process (Besomi et al, 2020), and as such this level of analysis is not accessible or affordable for most safety professionals; (ii) there are an abundance of back extensor muscles, many of which are not measurable with surface EMG, meaning EMG studies only capture a subset of back muscle behaviors; (iii) there is substantial inter-subject variability with respect to which back muscles individuals offload and in what proportion when exo assistance is provided (Lamers et al, 2020), meaning it is unclear which subset of muscles to measure or how to make equitable comparisons between different individuals or exos; (iv) EMG only captures changes in muscle activity, and fails to capture reductions in spinal ligament loading (e.g., during stooped bending); (v) the relationship between EMG and force is non-linear, and dependent on other factors like muscle length, contraction velocity, electrode placement, and body posture (Ranavolo et al, 2020;Trinler et al, 2018), which can confound exo vs. no exo comparisons if kinematics are not tracked or controlled (Koopman et al, 2020a;Lamers et al, 2020); (vi) there can be safety concerns with normalization procedures that require maximum contractions (Cholewicki et al, 2011); and (vii) there is within-and betweenparticipant variability when measuring EMG (e.g., due to motion artifacts, tissue conductivity) even without exos (Cholewicki et al, 2011), which can introduce reliability or interpretation issues, and these must be overcome with rigorous experimental design and data processing (Besomi et al, 2020).…”